GUI Bugfixes: 
change: revert vijava lib version to 5 (version 6 cause tcpip connection problem)
change: optimize get datacenter of VM, prevent get datacentername by VM chain via getParent()
bugfix: VM data will not more buffered, when start gui server (internal ManagedObjectNotFound will be throwns from XMLGenDom, when get parent of VM object)
change: ensure reset of the http request to release system resources
bugfix: (#18747) the first letter remain at drivegroup field, when remove complete datastore name
bugfix: buffer will not be refreshed, when remove task from taskgroup at tree
bugfix: (#18743) remove taskevent from schedule failed
bugfix: (#18639) state filter at media action table does not work
change: (#4676) Append final switch the skipRights flag and restart the GUI server
bugfix: (#18731) Correct switching of the absolute flag in restore events
bugfix: (#18729) VM config directory is empty, when restore VM config only
bugfix: (#18731) Correct radio button switching in migration tasks
bugfix: (#18672)  slotlist of loader will not be applied
change: set 'equal' instead of 'startwith' condition for filter vm by vcenter (dir vsphere <vcenter>)
bugfix: (#18071) remove duplicate dependencies
change: scan for 'class instance of VirtualController' + label contains 'SATA' to get SATA disk information, when create conf file for VM backup
change: set sep.version in poms to 4.4.2 to enforce new libs-snapshot-local in artifactory
bugfix: (#18724) Topology-tree willö be  closed automatically, when change client
bugfix: Avoid invalid values found error in TermsDaoTest
bugfix: (#18722) some clients of topology tree does not show name with sesam comment at tree, set sesam comment regex=^E0.* as show condition
new: filter cli 'list vsphere <vcenter>' by vcenter server
bugfix: test 'testAddTaskByTemplate' failed
change: (#4676) Authorization: Bind visibility of the permission management frame to new defaults entry "enable_permission_menu"; removed usage of sm_setup activate_uac; added usage of set_salt; persist root user with password via gui;
bugfix: (#17940) restore VM with SATA disk failed, reformat import vijava section with exclude lib to prevent duplicate classes
bugfix: (#17940) backup VM with sata VMDK write wrong type to VM conf file (VirtualController -> VirtualAHCIController)
bugfix: (#18713) Enable the start button if a task(_group) is selected
bugfix: (#18219) continue removal of nonExisting after last duplicate was found
change: read client id if not specified
bugfix: (#18713) Avoid NPE diuring immediate start of backup, if no task(_group) is selected
change: fix for client id; added support for colon
change: set clientId by create tasks
fix: correct logik in term calculator for weekenddays and weekdays
bugfix: corrected display on important messages button; corrected mapping of open messages in the notification center
bugfix: calculate next exec of yearly relative schedule with offset is incorrect, registry offset spinner at doCalculate listener
bugfix: (#18704) fix index detection in weekday
bugfix: (#18219) - CliMainIntegrationTest failed, fixed tasks insert
bugfix: (#18219) - tasks tests failed, fixed tasks insert
bugfix: (#18219) - task group is mandatory argument
change: (#18219) - added support for -N and -I
change: (#18219) - added -O check_existing and -g <task_group> support - added validator for mustache template
bugfix: (#18704) avoid infinite loop and fix detection os "last"
bugfix: (#18704) Correctly adjust week iteration and add unit test
bugfix: (#18704) adjust doCalculate in the termcalculator
bugfix: (#18704) Last Friday of month not calculated correctly
bugfix: (#17940) restore VM with SATA disk fixed (update vijava api to yavijava 6.0.01)
change: adjusted logging; cleanout code
change: add ssl thumbprint of current vcenter to command param '-a ssltp=<ssl thumbprint>', which will be used during VM operation for execute sbc command
bugfix: (#18625) persist the replication type format selected by the user
bugfix: (#18625) Fill the replication type selection menue with correct values
bugfix: linux vSphere server does not offer entry path 'VMware vSphere:' for browse vcenter, when use cli action 'dir'
bugfix: browse vCenter client with unreachable filesystem does not show at least vcenter path 'VMWare vSphere:'
new: (#18625) Replace migrationtask with task in the child rows of the migrations and replications table
bugfix: (#18686) existing taskgroup cannot be changed via GUI
bugfix: vcenter name appear twice, when browse VM directory with cli action 'dir'
change: improve error message "log file not found" at migrationresults dialog, show full file name
Revert "bugfix: open result log without existing file does not show missing file name at gui"
bugfix: open result log without existing file does not show missing file name at gui
new: (#18625) Equipped the replication type column with icons an short labels
bugfix: cli help of remove action sometimes appear several times
bugfix: add help example for cli add task with param -S "/<client>/<type>/<src>"
new: (#18625) introduced new icons for replications
change: improve log of vcenter access, always log vcenter password with 8 stars (,password=*******,---)
bugfix: run vm operation log password of vcenter unencrypted
bugfix: (#18636) 'PRT' with log will not be created for restore VM with attach/mount
bugfix: (#18677) option "mount saveset' is available for crypted savesets at restore wizard
bugfix: cli add task with arg of param 'S' starts with slash failed (/<client/<type>/<src>)
Revert "bugfix: cli add task with arg of param 'S' starts with slash failed (/<client/<type>/<src>)"
bugfix: cli add task with arg of param 'S' starts with slash failed (/<client/<type>/<src>)
bugfix: cli 'add task -S' cause wrong parsing, when source contain slash
new: cli param add task '-S'  <<client>/<backuptype/<source>', which generate task, no further param required
new: (#18625) Added a column for the replication type in the migration by status table
bugfix: (#18625) Correct property for replication type in the migration results info
bugfix: param '-t tasktype' is not visible at cli help task
bugfix: cli list vsphere does not show vcenter at VM path
new: (#18625) Add the replication type in the migration results info
bugfix: (#14471) Explicit check the lic file name when getting the license in a zipped file
bugfix: wrong error message, if add task via cli with non existing client
change: run command 'sm_arch' without param 'd <drivenum>', when delete datastore
change: (#18219) - added -O check_existing and -g <task_group> support - fix for -g
change: (#18219) - added -O check_existing and -g <task_group> support - returns success status on duplicate found
bugfix: (#18669) Avoid PSQLException relating to field hw_drives.throupghput (changed jdbcType from FLOAT to VARCHAR and use asyecific typeHandler)
bugfix: add task with group via cli will not be shown at gui (full DB refresh required)
bugfix: (#18673) create task with group with cli does not create group
bugfix: (#14471) Added the possibility to read a zipped lic-file in the licese dialog
change: instead of using invokelater have a swing timer that takes care of updating the data. this is easier to handle when clearing and seems to fix the "empty list" problem
bugfix: avoid multi-refresh in gui because of filter changed if it has not (make sure clientName=* equals to null)
change: some minor cleanup to allow easier debugging of empty views
bugfix: cli tests failed, which use taskparams, add null check for postObject
change: (#18219) - added -O check_existing and -g <task_group> support
change: get master server ip only once
change: fill client cache in a seperate thread and add status method
change: clean out duplicate overrides without extra functionality
 
SRC Bugfixes: 
bugfix: (#18366) ME10 Restore to generated RDB Diff,Incr. A zarafa restore did not restore the data.zbk file. Do not all BLOW_UP_SLL() again for fully derived SSL file otherwise the first sel line '{label} Subdir: ...' is skipped
bugfix: (#18366) ME10 Restore to generated RDB Diff,Incr. Last commit did corrupt other restore types. Now decide in PREPARE_SLL_FILE if LIS or LST must be used for further task and mail mode specific processing
change: Do not increase log level to 1 (ITEMS_VERBOSE_LEVEL) after sbc-3353 'Partial hardlink data retrieved: ...'
bugfix: (#18417) - ME13 Restore to generated RDB Copy,Full - Error - Error: DB Module: [ [CMailboxDatabase::DeleteFiles - Database::DeleteFiles:Could not find a part of the path 'C:\Program Files\Microsoft\Exchange Server\V15\Mailbox\SEP_RDB_20150603154226374'.]. Terminate the restore operation in case of occurring error in the vss_es.cpp;
BUGFIX: (#18707) SLU throws 'Loader not ready' after some changer movements. Set SCSI timeout to 600 for MOVE operations
bugfix: (#18366) ME10 Restore to generated RDB Diff,Incr - Error - ERROR: No DATABASE/USER line with offset. Now sm_lis_item writes XML data in GLIS '#' lines.
bugfix: (#18291) Generation restore of Exchange fails with 'xml directory for FULL saveset is missing'. Do not remove SSET metadata with xml data from GLIS file. Now these lines are added with first char '#' instead of '"'
bugfix: (#18702) Reset short name to empty string
bugfix: (#18738) Clarify the update dialog of the bsr install/update/deinstall/unchange.
bugfix: (#18702) System state restore ... Avoid messages 'sbc-3021: Info:  Item [\] already exists. Skipping overwrite...'
change: Restore of hard linked files counts every link to a file as processed hard link file
change: Avoid 'Unknown operating system (major:6,minor:3)' for WIN_2012_R2
bugfix: (#18702) System state restore failed with 'WIN32 API error: 5 - Access is denied'. In case of access violation during hardlink processing do not rename original file but restore to {file}.${pid}_new to avoid race condition with other processes
 bugfix: (#18702) System state restore ... Avoid setting short name if requested short name already set
bugfix: (#17222) Newline in file info.dat leads to garbled message in the GUI log. Line length increased to 1024
change: (#18031) reinserted the activate_uac switch for the activation of the user access control.
bugfix: (#18702) System state restore failed with 'WIN32 API error: 5 - Access is denied'. Avoid problem with race condition for recently restored hard linked files. Now if hardlink operation fails then check if files are already hard linked
change: Prototype for bMoveFileExW()
change: Obsolete if (psMTF_Global->bPipe) {...} removed
NEW: (#18730) 'i_setenv_sesam()': Try to add SMS bin directory to 'PATH' environment variable, so all SMS executables will be found. (no check for 'sm_sms_interface' in command call)
change: Do not set 'LD_LIBRARY_PATH' any more because Sesam programs will find shared libs by 'rpath' now
Bug 18727 - Browsing of Hyper-V single server returns empty VM list. Correction the previous commit SHA-1: 40e31b8b1bd28b0c340a9196fab49ae59ace5add
BUGFIX (#18727) - Browsing of Hyper-V single server returns empty VM list. Use the VSS API approach if there is no found any VM through WMI API.
bugfix: (#18716) Correct handling UNC file path has been implemented. Also VSS-writers status check and Windows services control features implementation has been added.
bugfix: (#18702) System state restore failed with 'WIN32 API error: 5 - Access is denied'. Next dir after processing shared hardlinks directory could be set to empty dir name during backup
new: (#18703) Windows 10. Step 1: Add os 'Windows 10' with SNAPSHOT-cmd and DISK_CHANGE-eject, mistyped 'Windows XP'
bugfix: (#17236) Startup sm_data_server failed with 'ERROR: DataServer: main: another sm_data_server is already running'. Set DEFAULT_LOCK_RETRIES 300 and add entry into sms.ini file [SMS_Server] Lock_Retries=300
new: (#18703) Windows 10. Step 1: Add os 'Windows 10' with SNAPSHOT-cmd and DISK_CHANGE-eject
change: Usage with verbose level
new: sm_sms_tapeutil now with function 'Read data loop'
change: compiler warnings resolved
BUGFIX: (#18721) sm_config_client adds access option -d to command which makes client configuration behind firewalls fail. Do not add option '-d' to ctrlc call.
bugfix: (#18719) sbc Unix timestamp files arn't updated. Since commit a32198c: 'Change XBSA FTP communication:' the file close was called even in case of XBSA transfer
change: Reset sMTFFileStatus.sFileInformation.dwFileAttributes only for WIN32
bugfix: (#18687) Hyper-V restore to SMB failed with 'Cannot find matching LIS line for item ...'
change: mistyped 'resturn'
bugfix: (#18702) System state restore failed with 'WIN32 API error: 5 - Access is denied'. Avoid to add data for already processed hardlinked files.
bugfix: (#18709) Case-insensitive file extension comparison has been implemented.
change: bugfix: (#18715) System state backup fails with [FindFirstFileW failed for ... 123 - ERROR_INVALID_NAME]. Obsolete 'Remember next dir' removed.
bugfix: (#18715) System state backup fails with [FindFirstFileW failed for ... 123 - ERROR_INVALID_NAME]. When backing up pending harlinked files for last volume the next directory must be remembered together with the length of the volume indicator
change: Browse information with hardlink [major:minor} must end with ','
bugfix: (#18705) A backslash removal procedure is continued until the last one is removed.
Bug (#18706) - System state backup fails with [BackupProcessing: Error: Could not found snapshot device]. The issue was caused by SHA-1: be025f1890ef4face6738a3a5b5a32e431d9af90.
bugfix: (#18684) Processing of "bin" and "vsv" files has been added to relocation correction algorithm.
NEW: Improved Sesam SAP HANA install scripts
new: (#18703) Windows 10. Step 1: Add os 'Windows 10' with major 10 and minor 0. Insert into table oper_systems
bugfix: use AppendBackslash instead of AppendBackslashW
bugfix: replace the concatenation of a backslash at the end with AppendBackslashW which checks if a backslash is present and only appends it if it is not present.
bugfix: Return again 1 for empty directories, e.g. C:/
bugfix: (#18697) System state backup with option 'hardlink_data_single' - Restore failed with invalid MTF stream ID. Restore hardlink meta data if file was processed
change: sm_client dir ... display hardlink major minor with [0X%08X:0X%08X] (similar to sbc hadlink meta file names
change: Added missing 'PDS_I_FILE_IN_USE_RENAMED' string in STR_ARRAY_ENUM_PDS
bugfix: (#18697) System state backup with option 'hardlink_data_single' - Restore failed with invalid MTF stream ID. Add missing SPAD stream after FILE DBLK
bugfix: Dovecot backup: Set function (backup/restore) correctly
BUGFIX: (#18463) - Bad blocks during system_recovery restore. VSS System_state: do not give same file name twice (with parent Directory in between)
add 'src/sbc_proxy/sesam_version.py' to gitignore
set sob version using hex
new SOB version matching Sesam kit version
bugfix: (#18694) Cleanup routine of Windows sbc at VMware backup is not called anymore, if STPD connection died
bugfix (#18420) - ME07 - Backup - error - sm_reformat_lis returned exit state 1. The logic error was present in the processing metadata file for further validation of databases files.
bugfix: Revert changes for 'ORACLE_SOB_on_XBSA' because not completely implemented. Merge with commit 2e81641 (master)
new: Set 'DB:restore_results.drive_num' to new drive if it has to be re-submiited into another queue
new: Target to copy vSphere mount helper scripts (re-scan SCSI, find NSS volumes..)
bugfix: (#18659) System state backup saves hardlinks on wrong path. Reset psMTF_Global->uiVolumeLength before processing partial saved hardlinks
change: MakeMTFDirName() now with formal parameter uiVolumeLength instead of global psMTF_Global->uiVolumeLength
bugfix:(#18678) - Path backup includes ASR Writer XML but excludes all components from ASR.
bugfix: (#18685) Total line is not written anymore when backing up via HTTP interface. BSAHttpEndData() was missing since commit ddac6da '#18655 BSR Pro Windows: Restore from DataStore with slow throughput 6 GB/h. Change XBSA FTP communication: BSAInit() establishes and BSATerminate() closes the FTP connection'
New branch to show EFI problems
bugfix: (#18683) sm_lis_items does not show the included VMs. In case of Hyper-V with CSV volume the VOLB line was parsed in wrong level, e.g. write level[5]*['Device=E:/' ... for 'Device=E:/qa-csv/Virtual Hard Disks,Volume=E:/qa-csv/Virtual Hard Disks,Machine=corefix1' v
bugfix: (#17917) - Special backup target 'system_reserved' to allow to backup the Windows 'system reserved partition'. The system reserved volume is snapped only.
change: Display only available (fixed and remote) volumes in case of restore mode: 'sm_client dir -o restore /'
* bugfix (#17917) - Special backup target 'system_reserved' to allow to backup the Windows 'system reserved partition'. Handle the source "system_reserved" in the vss module
bugfix (#17917) - Special backup target 'system_reserved' to allow to backup the Windows 'system reserved partition'. Handle the backup source "system_reserved" in the vss module.
bugfix: (#17917) - Special backup target 'system_reserved' to allow to backup the Windows 'system reserved partition'. Adapting the backup operation in the vss module by handling the new source "system_reserved" and added new function "GetSystemReservedVolume"
bugfix: (#17917) - Special backup target 'system_reserved' to allow to backup the Windows 'system reserved partition'. The system reserved volume is snapped only.
bugfix: Avoid calling xbsalog/xbsalogtrace/xbsatrace with arbitrary data given as format string
change: insert the build of the sbc_com_interface.exe in the build of the sbc
bugfix: (#18659) System state backup saves hardlinks on wrong path. USe the previous processed volume as parent directory for meta directory _SharedHardlinkData_
bugfix: (#18527) Incompatibility with C++11 implementation in MSVC2010 is fixed.
bugfix: (#18527) Network virtual switch absence situation is resolved, situation with impossibility to merge snapshot file system images on CVS is resolved also.
bugfix: Since commit a32198c: 'Change XBSA FTP communication: BSAInit() establishes and BSATerminate() closes the FTP connection' the backup to null device or local file stalled in BSAInit. Now the BSA... functions are only called if backup with XBSA is requested.
change: Indentation revised
change: Avoid lock after NOOP. xbsa_def.h
change: Avoid lock after NOOP
change: Set return to BSA_RC_SUCCESS if 553 with WSAECONNRESET was received
change: Continue with reading after 200 'NOOP' Command successful.
change: BSAEndData() must be called after BSAGetData() and BSASendData() to close the SDATA socket
change: Allow repeated call for all BSA... functions
change: BSAGetObject() Allow repeated call
bugfix: (#18664) some VMs were not shown when browsing a hypercluster because the fflush of the stderr happened too late. The fflush was move about the printing of the STATUS message.
change: Version string without CVS Id
bugfix: (#18655) BSR Pro Windows: Restore from DataStore with slow throughput 6 GB/h. Reset cpCode before processing BSAGetObject()
bugfix: (#18643) vSphere backup fails with warning: DB Module: [sbc_vmware error]
change: Recent change sets user to 'SESAM_SECURE_AUTHENTICATION' with keyword '' but PWD file does not accept empt value
change: Use 'SESAM_SECURE_AUTHENTICATION' for STPD user access
change: use the new O&O installer version 10.0.69
change: BSR Pro Version 10.0 uses a different product key
change: New defines for socket shutdown() flags
bugfix: (#18655) BSR Pro Windows: Restore from DataStore with slow throughput 6 GB/h. Change XBSA FTP communication: BSAInit() establishes the FTP connection and BSATerminate() closes the FTP connection.
bugfix: (#18652) insert the bsr_pro=none option in the sm_setup install/update routine.
bugfix: remove the bCBtFlag from the parameter list of the SetArchivePipe.
bugfix: inserted the packdata in the clean rule.
bugfix: inserted the basics.dll in the clean rule.
bugfix: remove the quotes from the product code and the installer package name
bugfix: (#18652) due to a major version change in the O&O installer an additional uninstall chained package was inserted. And a different registry key is now checked for O&O. From now on in case of a major version change of O&O an additional uninstall routine has to be added and the new product code has to be inserted.
bugfix: (#18656) sm_sesam.exe cores when opening properties of failed vSphere backups. Stack size increased to 4GB
change: use new O&O version 10.0.68
bugfix: revert all left over VMDK_CBT specific changes.
bugfix: (#18551) Restore a split saveset from HARD_DISK runs endless with error MTF_ERROR_STREAM_NOT_FOUND. Avoid move operations on following disk
BUGFIX: (#18404) (from branch v_4_4_2_57_branch gitid e929af0d) VMware CBT backup with Windows data mover slows down when backing up small blocks. In case of CBT backup do not check for hard linked file
change: Do not build sqlite DLL with common build. Use rule in su/src/makefile.win32 for explicite sqlite_lib build together with rule all
bugfix: (#18643) vSphere backup fails with warning: DB Module: [sbc_vmware error](reverted skip zero block changes)
change: If called for restore with option '-l list' then create script to validate given saveset
bugfix: (#18643) vSphere backup fails with warning: DB Module: [sbc_vmware error](dont check pipe flag during restore)
NEW: (#18429) BSR Windows: O&O DiskImage Backup and Disaster Restore - Seek in saveset is to slow. Seek with offset less than block size 64k was not implemented
change: trace message revised
bugfix: (#18652) insert more log lines and insert the REGDB_OPTION_WOW64_64KEY option
bugfix: (#18652) the automatic replacement of the O&O version in the build script did not work properly.
bugfix: (#18643) vSphere backup fails with warning: DB Module: [sbc_vmware error](skip blocks with zero size)
change: use the O&O DiskImage subkey to obtain the installed O&O version. Compare the given O&O version with the one included in the installer.
change: removed the check for the O&O version number since it is not needed anymore. This would only be necessary if O&O would be updated but since a deinstallation takes place no version number is needed.
change: Reset multiply factor for offset
bugfix: (#18643) vSphere backup fails with warning: DB Module: [sbc_vmware error](resolved merge errors)
change: Additional result codes
change: Set registry keys to version 10 specific values
bugfix: (#18643) vSphere backup fails with warning: DB Module: [sbc_vmware error]
change: partially revert commit 97ff08d631c 'bugfix: crash oh Hyper-V relocation restore'
bugfix: Insert the missing parameter for the ArchivePipe function in the recreate_pipe function
bugfix: remove the ifdef PIPE_USAGE from the Read_STAN_STREAM method.
change: inserted the new O&O installer version 10.0.64
change: find_com_intf(): Trace the sbc_com_interface command
bugfix: hyper-V relocation restore
bugfix: crash oh Hyper-V relocation restore
bugfix: (#18640) BSR Windows: Restore data connection ends with WSAECONNRESET (10054) Connection reset by peer. Set SO_LINGER with 10s to allow a graceful shutdown of STPD data port when closesocket() is called
remove '#include <stropts.h>', because probaly not necessary at least for LINUX and Solaris
Revert "Adoption for Solaris"
Revert "change: fix online check after SUNOS fix"
BUGFIX: (#18404) VMware CBT backup with Windows data mover slows down when backing up small blocks. In case of CBT backup do not check for hard linked file
change: Call GetAlignedBuffer(size)
BUGFIX: (#18404) VMware CBT backup with Windows data mover slows down when backing up small blocks. In case of CBT the backup data is transferred via pipe and the device check must be skipped.
Revert "bugfix: (#18546) Net backup with sbc on Solaris 11 doesn't work (SMS also need to be fixed)"
BUGFIX: (#18147) system_state backup should notify AD /repadmin after successful backup. Allow to execute System REcovery job with level FULL to truncate logs of AD (NTDS).
change: Due to recent merge the necessary #include "gnu_regex/regex.h" was missing in cm_common.c
bugfix:(#18631) Excluded unnecessary VSS-writer calls in case of VM restoration with relocation along all the algorithm.
bugfix: (#18489) DO NOT REVERT IT! Comment "_BUGFIX_18489" in the "sbc_w32_os2\vss\vss_server.h" instead.
Revert "bugfix: (#18404) VMware CBT backup with Windows data mover slows down when backing up small blocks"
Revert "bugfix: fixed build on linux packdata"
Revert "bugfix: vfs on windows"
Revert "bugfix: fixed sbcmsql backup"
Revert "bugfix: hyper-V relocation restore"
Revert "bugfix: crash oh Hyper-V relocation restore"
Revert "bugfix: (#18404) VMware CBT backup with Windows data mover slows down when backing up small blocks  (use one thread for all cbt blocks)"
Revert "bugfix: close buffer correctly if reading from disk failed"
Revert "bugfix: (#18404) VMware CBT backup with Windows data mover slows down when backing up small blocks (redesigned cbt backup)"
Revert "bugfix: (#18404) VMware CBT backup with Windows data mover slows down when backing up small blocks (made additional cbt adjustments)"
Revert "bugfix: (#18604) VMware configuration backup cannot remove temp directory anymore"
Revert "bugfix: (#18614) sbc cores at VMware CBT backup"
Revert "bugfix: (#18615) Restore of VMware host works, but sbc cores at the end"
Revert "bugfix: fixed mssql restore"
Revert "bugfix: fixed vmware CBT restore"
Revert "bugfix: write buffer end buffer only in the end vmdk transfer"
Revert "bugfix: adopted new cbt backup for Linux"
Revert "bugfix: (#18615) Restore of VMware host works, but sbc cores at the end"
new: Print out GitID
bugfix: Avoid overflow in block number if called with big offset
change: Revert to #include 'regex.h' (after recent change in cm_common.h)
change: Avoid conflicting #include <gnu_regex/regex.h> in cm_common.h
change: Compile with #include <gnu_regex/regex.h>
change: Avoid conflictzing regexp types with (gnu regex.h <> local regex.h)
change: Revert changes in common: cm_kernel.c, common.h, common.c, common/... except the cm_common.c get_addr_len() function
add 'gnu_regex.h' to build on 4.4.2
change: fix online check after SUNOS fix
bugfix: (#17383) implemented topology support in sm_auto_conf_hw
BUGFIX: (#18498) Restart (stop and start) 'sm_sms_watch' during drive configuration to recognize changed drive configuration
bugfix: (#18615) Restore of VMware host works, but sbc cores at the end
bugfix: adopted new cbt backup for Linux
bugfix: write buffer end buffer only in the end vmdk transfer
bugfix: fixed vmware CBT restore
bugfix: fixed mssql restore
bugfix: windows build issue
bugfix: (#18615) Restore of VMware host works, but sbc cores at the end
bugfix: (#18463) Bad blocks during system_recovery restore. PACKDATA: If already processed directory appears again then process it again
bugfix: (#18614) sbc cores at VMware CBT backup
bugfix: (#18544) SBC trys to restore files with double backslash in path during system state restore. Recent changes revised.
new: Use 'AR' macro for 'ar' so it can be overwritten at one place. (ar -X64 on AIX for example)
Bug (#18544) - SBC trys to restore files with double backslash in path during system state restore. The issue is caused by the commit 0c897ac5cf5add3a80052a528d9ff902b531bbed
bugfix: (#18604) VMware configuration backup cannot remove temp directory anymore
bugfix: (#18404) VMware CBT backup with Windows data mover slows down when backing up small blocks (made additional cbt adjustments)
bugfix: (#18404) VMware CBT backup with Windows data mover slows down when backing up small blocks (redesigned cbt backup)
bugfix: close buffer correctly if reading from disk failed
bugfix: (#18404) VMware CBT backup with Windows data mover slows down when backing up small blocks  (use one thread for all cbt blocks)
bugfix: hyper-V relocation restore
bugfix: crash oh Hyper-V relocation restore
bugfix: fixed sbcmsql backup
add '<sesamdev>/include'
bugfix: vfs on windows
bugfix: fixed build on linux packdata
bugfix: insert the $(PCRE_DIR)\$(PLATFORM) in the makefile.
change: Move 'i_pcre_find' and 'pscanf' from 'cm_kernel' to 'cm_common'
bugfix: (#18404) VMware CBT backup with Windows data mover slows down when backing up small blocks
bugfix: fixed wrong path while restoring. Added backslash to separate file name and folder name if missed
BUGFIX: (#18578) SMS read tape with slow throughput. Set read buffer size to 64k and tape block size 64K if sms.ini [SMS_Tape] Read_Block_Size=64K is set
new: Use 'statvfs64' on AIX
bugfix: Large file support for AIX. (-D _LARGE_FILES)
use define 'UNIX' instead of '_UNIX'
bugfix: (#18578) SMS read tape with slow throughput. If sms.ini [SMS_Tape] Read_Block_Size=64K is set then set uiMaxBlockSize = BLOCK_SIZE_64K to set the buffer read size to 64K too.
BUGFIX: (#18605) Oracle backup fails with 10054. RecvN: In case of partial read remember that EOF was already read and return EOF if next recv fails.
Support a platforms that doesn't have the 'C99' standard 'uint*_t'.
bugfix: (#18546) Net backup with sbc on Solaris 11 doesn't work (SMS also need to be fixed)
Adoption for Solaris
'PTHREAD_MUTEX_RECURSIVE_NP' is available on Linux only
Adoptions for 'AIX'
Using 'uint8_t' insread 'u_int8_t' because it doesn't supported under Solaris.
new: Avoid using GNUTLS on unix systems (remove libs, needed only for gnutls)
new: Avoid using GNUTLS on unix systems
change generation of 'sesam_version' so it doesn't relay on 'bash'
change: In case of COM session rename log file to stpd_[pid].com.log after closing the log file
bugfix: (#18578) SMS read tape with slow throughput. Allow to set sms.ini [SMS_Tape] Read_Block_Size=64K to ignore variable block size (0)
change: Add message in case of error '2008: Invalid Block Size.' to trace the current block size
Revert "change: use the new O&O installer package version 10.0.61"
bugfix: (#18621) use the entries of the sm.ini instead of the registry values to resolve the path in the file_copy function. This is needed since the needed registry entries are not present on the RDS.
change: use the new O&O installer package version 10.0.61
new: merge from master: Quote path names
bugfix: Add 'sm_set_java' to client programs
change generation of 'sesam_version' so it doesn't relay on 'bash'
change: use XCode 4.5 instead of 3.1 for the build.
change: switch to the current SDK version of the macdev1082 for the mac build.
bugfix: (#18558) RPEX: Call 'sm_lis_items' with 'gen' flag to get EDB path from full backup. Do not throw exception, if EDB or LOG path aren't found in LIS and start RPEX with top level mount path instead
BUGFIX: sbc_lvm: 18573: strip ending / of path so mount command can find correct volume
bugfix: (#18573): Backup over LVM mount: Strip trailing '/' from source for searching mount point, because if 'soucre=mount point' it will not be found.
BUGFIX: (#18476) Truncate saveset id from migrated saveset to 16 chars because 'backint' accepts only 16 chars
bugfix: Fix SQL syntax error during creation of DB triggers. The error were not fatal
change: sob_com_intf(): Trace for read bytes in loop only if log level is 3
change: Only trace the answer buffer if called with log level 3
BUGFIX: (#18553) Un-mount saveset (stop 'sm_vfs') on RDS and not on Sesam server
bugfix: (#18560) Windows ORACLE backup failed in version 4.4.2.57. Adapt recent changes in SOB and XBSA for Windows
change: In case of Windows the program file name must be enclosed in "
change: find_com_intf(): In case of x86 retry with RegOpenKeyEx(...,KEY_WOW64_64KEY)
bugfix: (#18496) inserted the move of the python specific zip file for sm_sshd and changed the python33 dll move to python34 dll.
change: Version information adapted to: 'sbc-3036: Info: # SEP XBSA, VERSION: %s%s Build: %s, Released: %s #'
BUGFIX: (#15635) Use results.based_on and .based_on_full to avoid wrong FULL/DIFF/INCR saveset chains. Use only savesets which are referenced by results.based_on and .based_on_full. New switches: -o metadata|force
bugfix: (#18464) - Microsoft group policies aren't saved by "system_recovery". SYSVOL files are backing up firstly.
bugfix: (#18532) BSR Pro Windows: Better error messages instead of 'Error: Operation failed:'. Now sbc_bsr.py adds the error code and message ID to the 'Operation failed: ...'
bugfix: (#18531) BSR Pro Windows: Fisrt backup failed with sbr-3500: Info: BSR PRO initialize from PWD file returned (0XC0000034) OBJECT_NAME_NOT_FOUND. Check first if the PWD file exists.
new: (#18529) BSR Windows - Client browser must allow to select BSR Windows for BSR Pro. sm_client dir / now displays '/BSR Windows:' together with information 'BSR Pro Version 9.5.65'
new: 16893 - Skip VSS Meta Data xml backup documents from Restore Wizard Select Tree. Step 2: All lines in LIS between 'SSET metadata' and 'ESET metadata' are skipped in sorted LST and Level Sorted LSL file. Step 3: Set sm.ini [Params] lis_with_metadata=1 to reactivate old with metadata in LST/LSL
new: 16893 - Skip VSS Meta Data xml backup documents from Restore Wizard Select Tree. Step 1: Indicate 'SSET metadata' and 'ESET metadata' in LIS
bugfix: (#18420) ME07 - Backup - error - sm_reformat_lis returned exit state 1. If meta file is missing in case of old sbc < 4.4.2 then skip meta file processing
change: Add sm_exchange.py to dependency rule for sm_reformat_lis
change: XX_AUTO_LOGIN_INCORRECT with 'Please adjust your TCP/IP name resolution.'
bugfix: (#18507) - System state backup fails with error VSS_E_OBJECT_NOT_FOUND. Move the trace with "SetBackupSucceeded" up.
bugfix: (#18507) - System state backup fails with error VSS_E_OBJECT_NOT_FOUND. Skip calling "SetBackupSucceeded" for ASR writer as it could lead to VSS_E_OBJECT_NOT_FOUND error in some cases
bugfix (#18507): System state backup fails with error VSS_E_OBJECT_NOT_FOUND. Skip custom writers for SetBackupSucceeded(), added trace
Bug 18238 - HyperV backup hangs at DllMain: Uninitializing COM ... until newday - use CoInitializeEx instead of CoInitialize, upgraded/fixed decorator for IVssBackupComponents object to get object released after operation is finished.
Revert "Revert "* bugfix (#18487) - Simultaneously started VSS backups fail with VSS_E_SNAPSHOT_SET_IN_PROGRESS - simple file backup is occurred with "VSS API error: CProcessesSynchronizer::SharedDataAccess: Failed to lock."""
Revert "Revert "bugfix (#18487) - Simultaneously started VSS backups fail with VSS_E_SNAPSHOT_SET_IN_PROGRESS - improved the mechanism of simultaneously started VSS backups for different solution.""
change: Call x_DumpEx() for read block only in log level 5
NEW: (#18429) BSR Windows: O&O DiskImage Backup and Disaster Restore - Seek in saveset is to slow. Seek to given offset instead of read till offset is reached
change: Compiler warnings resolved
Revert "bugfix (#18487) - Simultaneously started VSS backups fail with VSS_E_SNAPSHOT_SET_IN_PROGRESS - improved the mechanism of simultaneously started VSS backups for different solution."
Revert "* bugfix (#18487) - Simultaneously started VSS backups fail with VSS_E_SNAPSHOT_SET_IN_PROGRESS - simple file backup is occurred with "VSS API error: CProcessesSynchronizer::SharedDataAccess: Failed to lock.""
bugfix: Triggers for restore_results, migration_results and media_results have to reference the specific primary key columns
bugfix: (#18503) vSphere restore fails with VM exception caused by a incorrect datastore path. MTF_END_PROCESSING must have a negative value.
bugfix: (#18503) vSphere restore fails with VM exception caused by a incorrect datastore path. Device delimiter is set to '/' since commit 5f4520a64d: (#18489) Hyper-V selective restore of one vm from a Hyper-V single server fails
change: Renamed  MTF_ERROR_END_PROCESSING to MTF_END_PROCESSING (not an error)
change: Trace message with format specifier for long: 'Unexpected: Different BSA handles session[%ld] %ld <> argument %ld'
change: Renamed  MTF_ERROR_END_PROCESSING to MTF_END_PROCESSING (not an error)
change: include the new O&O version 9.5.65
sbc_jira: fix removal of created backup dump file
* bugfix (#18487) - Simultaneously started VSS backups fail with VSS_E_SNAPSHOT_SET_IN_PROGRESS - simple file backup is occurred with "VSS API error: CProcessesSynchronizer::SharedDataAccess: Failed to lock."
bugfix (#18487) - Simultaneously started VSS backups fail with VSS_E_SNAPSHOT_SET_IN_PROGRESS - improved the mechanism of simultaneously started VSS backups for different solution.
fix: (#17367) STATS command didn't work in case of FULL backup
new: (#17367) Implement stats command when backing up MS-SQL database
bugfix: (#18386) - Show included/excluded VSS writers in log level 0 - Change the sbc trace to sbc message 3500 for excluded writers
change: (#18453) introduced 4 bit version field in extended attributes header; fixed a bug for encryption type
new: Added the mtime colum triggers for restore_results, migration_results and media_results
bugfix: (#18486) sbc cores on http and https backups since V4.4.2.55
change: (#18480) Sepuler is offline after system reboot. Reset counter after successful READ_INI()
bugfix:(#18483) - false error message during Hype-V cluster VM restoration.
bugfix: (#18489) Hyper-V selective restore of one vm from a Hyper-V single server fails. Turn device identifier delimiter '\' into '/' to allow a match with the Hyper-V item path with sm_reformat_lis
change: Last commit added browsing for 'System_Recovery' but there should only be the critical volumes, not all volumes displayed, so avoid this extra browsing level again.
new: (#18490) Insert a list of VSS writers during backup. Client browser now displays 'VSS:/{writers}'
change: sFileInformation not available on UNIX
change: Reset iTimeCnt
BUGFIX: (#18475) SAP / ORACLE restores only 200 of 300 Files. Avoid problem whit wrong queue entry id by using uniq job id
new: insert the sbc_vss.exe in the client and the gui package also.
bugfix: (#18480) Sepuler is offline after system reboot. Probably a read sm.ini file problem. Now retry up to 12 times with delay 5s.
change: Added '-EXPORT:x_Snprintf' to LINK_EXPORT
bugfix: use a string instead of an integer to find out if the drive group is encryption_capable.
change: x_DumpEx(): Instead of printing the address the size in hex is added to the BLOCK INFO. This allows to compare different dumps without getting different BLOCK INFO lines
bugfix: (#18474) BSR Windows: O&O DiskImage Restore failed with core dump in BSAInit(). Check Session index in BSAInit() and reset session index in BSATerminate() to avoid array index overrun
change: switched to the new O&O version 9.5.60.
bugfix: (#18470) The same file is stored endless during path backup. Reset sFileInformation if called for Harlink group processing
bugfix: (#18464) - Restore of microsoft group policys doesn't work with system_recovery - added "DFS Replication writer" into system_state backup, making the general solution for involving system writers starting from Win2008
change: Copy sbc_vss.exe to SEP sesam GUI and Client packages
bugfix: (#18057): implemented missing void functions
add log if waiting for EOM semaphore
change: Compielr warnings since commit d2af0c7 resolved
change: Structure for semaphore ctrl_restore'dn' with same name [g]semCtrlRestore in all modules
change: Code revised: Extra function DELETE_EVENT_FROM_DB() to delete obsolete ..._events and log file name changed into 'sm_sepul_event_{schedule}-{ev_type}-{object}_{hhmmss}_{sesam_date}.log
new: Add backup type 'JIRA'
new: Add GitID and check arguments
change: Recent changes fetched from master (commit 8f723abd): -F none default for restore, use r.json(), parse 'JIRA:/'
change: Put 'replication' into a sub-function
change: use VDDK 5.5.4 instead of 5.5.3
new: Flush log message to STDERR
new: Create symlink to 'sesam_version.py'
change: (#18418) Call 'sbc_vadp_nss_post' script before umount of VMDK
new: (#18419) umounting NSS pools with special characters in the name
bugfix: (#18455) - ME10 - Backup - error - Error: VSS API error: CVssServer::CreateSnapshot: No volume could be determined. Side effect of the commit 399e859f304c921ccf93ed5b1e64a5b3e73d345d
change: Adapted initializer for static _MTF_GLOBAL sMTFContext due to recent changes
BUGFIX: (#16472) Backup archive shortcuts (Metalogix HSM Archive). Set MTF Common Block Displayable Size to store file size of original HSM file and to set it in case of restore
bugfix: (#18430) restore from a datastore over http and https fails with no "no tape in drive"
change: (#18451) MS SQL is able to recover into new DB using sm_restore -a relocate_log_files (automated Logical Files relocation according to data in LIS file)
change: #define KEY_LEN 128 added in xutil.h
change: Avoid truncation of semaphore file name. Now use SCT_SEMA char szSema[KEY_LEN] (128). All semaphore caller must be recompiled.
BUGFIX: (#18449) sbc_rear: end backup with warning if debian jessie is detected
change: Output of 'Parameter file' /'Input file' with colon ':'
change: SSB: Trace input file name
bugfix: (#18436) SAP Restore ends with: XBSA Call BSAGetObject failed with message: Access to the requested object is not possible. Add output to trace the parameter file name.
BUGFIX: (#17649) notes module does not backup anything if backup source ends with /. Remove '/' if given as last char of backup source.
Fix recommendation for maxPages.
WORKAROUND: (#18434) Wait 60 sec. to give 'sm_sms_watch' enough time to occupy EOM semaphore and avoid dead lock
change: Use 'long long' for 'DB:terms.id/*_events.id' to soter unique id
bugfix: (#18430) restore from a datastore over http and https fails with no "no tape in drive"(applied workaround)
bugfix: initialize variable work_dir, so it works for unix also.
change: 'DB:terms/*_events.id' as BIGINT
BUGFIX: Do not set EOL back to an earlier point in time, if EOL of pool used for DIFF/INCR backups is shorter then of FULL. (fix: SQL 'or' clause not in parentheses
bugfix: (#16472) Backup archive shortcuts (Metalogix Archive). Avoid Warning with IO_REPARSE_TAG_HSAG_HSM 0X25
BUGFIX: (#18097) the working directory of the file to execute was not set correctly, hence the file was not found during execution. Now the working directory is explicitly set when executing the the command via Popen
BUGFIX: (#18436) SAP Restore ends with: XBSA Call BSAGetObject failed with message: Access to the requested object is not possible. 553 RETR Failed. Das System kann die angegebene Datei nicht finden. In case of a migrated saveset the saveset name contains an additional '@' delimiter.
change: Semaphore ctrl_restore'Dn' was never occupied
bugfix: VMWare Mount: Separate drive letters by ','
new: added some missing third party licenses to the java sources license list.
new: added getdelim/getline licence for the C sources.
bugfix: (#18397) - system state backup fails with error: Could not found snapshot device. Fix the appearing error messages in the Windows Event Log
change: sm_vfs accepts -f option
sbc_jira: import requests module in class only
bugfix: erased a core. In case the select column and the key word 'from' were not separated.
bugfix: (#18397) - system state backup fails with error: Could not found snapshot device. Changed the logic of involving system-critical components according to https://msdn.microsoft.com/en-us/library/aa384630%28VS.85%29.aspx?f=255&MSPPError=-2147217396
bugfix: (#18409) Single item recover: Call client side rescan scripts for action 'detach-ro' (not onyl for 'detach')
bugfix: (#18424) WINDOWS: GUI browsing of attached vmdk fails due to wrong output of 'sm_vmdk_interface mount' command. The common sm_popen() fucntion must turn exit state None into 0 for Windows too.
change: Avoid log line 'SEL Restore or single selection (offset+size)' in log level 3
change: Trace message revised: 'Return code 0X%lX so skip remove ...'
change: Windows: Use #undef snprintf and #undef vsnprintf before redefine to secure version, e.g. vsnprintf_s(A,B,_TRUNCATE,C,D)
change: Obsolete lines removed
change: Give file name with ASCII
bugfix: Comment out 'requests' module for JIRA, because 'sbc_proxy' could not be executed else
bugfix: (#18406) sm_restore cores when trying to restore a vSphere backup to other vCenter or ESX. The string length for the command must be increased to MAX_CMD_LEN
change: (#18325) BSR Windows: O&O DiskImage Backup and Disaster Restore - Generate Network Server settings via sbc_bsr. Check if values based on Sesam Server and COM port are yet stored in BSR PWD file and add if necessary
new: inserted the new O&O installer version 9.5.54
NEW: (#17198) Start external restores from data stores with queue allocation '_READ_' so they can run parallel.
new: inserted the oodipwdsec.dll in the installshield project.
change: In case of license violation add the message about the violation to the mail subject
change: reinserted the build of the 5.1.1 vddk for x86 build only.
change: v_set_log_main() Setting structure for trace() revised
bugfix: (#18361) slu on win only; set IO flag to out when using SPOUT command
new: inserted the oodipwdsec.dll in the server, client and gui kit. And inserted the newest version of the O&O kit in the build.
sbc_jira: add include path in makefile
change: Trace message revised ('gv_enable_tape_crypt' not longer used)
change: v_set_log_main() revised: Call first WRITE_TRACE() after checking SM_INI [Params] gv_log_milliseconds=YES. Check language only if called for the first time
change: Compiler warnings resolved
bugfix: (#18398) sm_restore cores when trying to restore a fdi chain (path) from incr backup. Avoid string overrun in i_INSERT_INTO_RESTORE_RESULTS()
change: Compiler warnings resolved
bugfix: (#18398) sm_restore cores when trying to restore a fdi chain (path) from incr backup. Avoid string overrun in i_INSERT_INTO_RESTORE_RESULTS()
change: Code revised: obsolete strlcpy() removed
bugfix: (#18395) Windows sm_sepuler ends with STATUS_STACK_BUFFER_OVERRUN. Avoid string overflow in char szDbCmd [...] in function iGET_TERMS()
change: x_Snprintf() Comment revised
bugfix: (#18395) Windows sm_sepuler ends with STATUS_STACK_BUFFER_OVERRUN. Avoid string overflow in char szDbCmd [...] in function iGET_TERMS()
change: x_Snprintf() Comment revised
bugfix: (#18361) use correct buffer size when sending SPOUT command
bugfix: (#18361) The clear of the encryption will only be executed if the encryption is on.
change: erased the copy of the vmware_vddk5.1.1.zip file since it does not exist anymore.
bugfix: Resolved compile error C2365: 'log' : redefinition; previous definition was 'function'. Rename log to log_func
CHANGE: Windows: Use CL compile option /Ox (Full Optimization) to increase performance
BUGFIX: (#18388) Every day a '0-days-warning' license warning appeared without ending the license. Avoid to extend the license timeout
bugfix: change the old database values in the media and media_pools table for encryption from n/y to 0/1 since these values were changed over time.
change: for the sm_remote_installer and the sm_main in case a password is required for the remote installation or to run the service using a different user the password is now checked using the getpass function if the password is not given using the -p parameter.
sbc_jira: check_zipfile
sbc_jira: check processlist for possible jira instance
sbc_jira: browse
change: added sense codes 0xE/0x0,0xE/0x1,0xE/0x2,0xE/0x3
change: only report sense key specific information if sksv bit is set
change: New 'I002-LICENSE All license issues have been resolved.'
BUGFIX: (#18388) Every day a '0-days-warning' license warning appeared without ending the license. Now check if reduced date is invalid
change: sm_info hash if compiled with /DWITH_HASH
change: Windows: tmpfile() tries to create file under C:/ so EACCES may be returned. Use GetTempPath() and _tempnam()  to create unique file name instead.
change: typo fix
change: Messages with level 4000 (Trace) must only appear if log level >= 3 (SBC_TRACE_LEVEL)
change: missing argument in logs
BUGFIX: (#18097) The remote execution of powershell scripts using sm_ssh failed with client version 4.4.1.48. Solution: Add the extension ps1 to the command. Also set the environment variables for sesam.
NEW: (#16046) License Overview: The Clients which are not listed as 'Backup Clients' must be shown too. Now 'Clients' are displayed in addition to 'Backup Clients'
bugfix: Set 'DB:results.fdi_type' to 'G.' if a task group is not started because of missed deadline. TODO: Write a result entry for every task in group so they are offered in restart dialog
bugfix: (#18361) in case slu -E clear returns error during sm_config_drives and a tape is loaded the drive is marked as not encryption capable.
NEW: (#16112) Exchange User based license. Retrieve Exchange User mailbox counter only in case of full show 'sm_info C' (e.g. after day change) otherwise use stored used value
change: gethostbyname() trace lines revised
change: remove the build of the vddk 5.1.1. zip file from the kit build. Hence the vmware_vddk5.1.1.zip will not be included in the kit anymore.
bugfix: If gethostbyname() failed then struct hostent *spHostEnt must not be used
bugfix: (#18376) When executing drive_handler mount it is checked if encryption has to be set. In case the given device is not capable of encryption the drive handler returned error. The return value is now changed to SM_OK in case of a non encryption capable device.
bugfix: SQL Error in iCHECK_SPARE_MEDIA() since commit 118bca0: Bugfix (#17564) Media init has used medium from SPARE pool even it has been locked by user
change: slu fix for Data encryption management capabilities page
sbc_jira: check created dump file for contents
change: Warning message about SimpleRecovery Model does not display database name since commit a57d32cc. Check if item is given with wchar
bugfix: (#18338) Cristie BMR restore fails with SMS ERROR 0x4000000E : SMERR_DEVICE_HAS_BLANK_MEDIA. Do not set 'size=...' in case of CBMR
change: Warning message about SimpleRecovery Model does not display database name since commit a57d32cc6: 'bugfix: (#17320) MSSQL Backup of database with special characters fails. Use wide char to pass database name to SQL API'
bugfix(#15899): URI/URL encode backup source(added basic function for parsing and checking)
sbc_jira: help output, loglevel
sbc_jira: @@file not needed
sbc_jira: pass generated filename to imort script
bugfix: (#18365) Disaster Recovery of SESAM_BACKUP does not work in 4.4.2. Only a BSR_WINDOWS / BSR PRO restore with level FULL may need segment file to read data from specific positions (seek offset)
BUGFIX: (#18368) Insert into 'DB:cal_sheets': Use 'DB:terms.id' to create unique primary key. Before a variable has been incremented for every schedule, but that doesn't work if 'sm_sepuler_cal' is called for a single schedule
bugfix: (#18365) Disaster Recovery of SESAM_BACKUP does not work in 4.4.2. Added sizeof(szBuf) to i_CHECK_POOL_AND_EOL()
bugfix: (#18365) Disaster Recovery of SESAM_BACKUP does not work in 4.4.2. If backup type is 'Path' and restore level is 'f'ull then a missing segment file error is ignroed to allow a disaster recovery of a SESAM_BACKUP saveset
NEW: Add new task type 'JIRA'
bugfix: shift SetBackupSucceeded() and BackupComplete() functions and call them after the CheckSelectedWritersStatus()
sbc_jira: create filename on restore
change: missing argument
change: test unit ready - test 3 times before giving up
change: (#18325) BSR Windows: O&O DiskImage Backup and Disaster Restore - Generate Network Server settings via sbc_bsr. 1st draft.
NEW: (#16112) Exchange User based license. First skeleton implementation to get the Exchange Server mailbox user count (DRAFT)
bugfix: set the variable bCartridgePresent which tells if the cartridge is loaded or not. Only if the cartridge is loaded the clear and set encryption will be executed.
change: EXPORT_WHOLE_DB: Avoid to check 'drop vies ...' lines
change: #include strlcpy.h
change: vDoLog() Use atomar write() instead of fwrite() to write into log file
bugfix: (#18332) - MS SQL 2012 backup is running endless. Handled the VD_E_ABORT returned value of IClientVirtualDevice_GetCommand.
Revert "change: erased the check for the global variable gv_enable_tape_crypt hence LTO encryption is now activated."
bugfix: sm_info c UPDATE_TIMEOUT_DATE returned truncated date, e.g. 2016052
bugfix: CHECK_BACKUP_LICENSE() failed for not license free backup types, e.g. 'SAP'. New: License output for 'not' Backup Clients
bugfix: (#18312) sbc -f <sourcefile> works only for the first 256 entries the source file -> state: success
change: i_LIC_COUNT_CLIENTS() Implement indentation blocks witn '{'/'}' for backup_types and tasks (unused)
bugfix: (#18356) Notes Backup ends with Warnings due to extra DB_ItemFinishStatus(). Skip extra DB_ItemFinishStatus() call
change: Call i_CHECK_LICENSE_ALL() with string buffer and declare LIC_DATA char szMsg[MESSAGE_LEN]
change: Running sm_backup on CLI must insert result entry and return correct results.cnt
bugfix: CHECK_BACKUP_LICENSE() failed for not license free backup types, e.g. 'SAP'. New: License output for 'not' Backup Clients
change: erased the check for the global variable gv_enable_tape_crypt hence LTO encryption is now activated.
bugfix: (#18312) sbc -f <sourcefile> works only for the first 256 entries the source file -> state: success
change: i_LIC_COUNT_CLIENTS() Implement indentation blocks witn '{'/'}' for backup_types and tasks (unused)
bugfix: (#18356) Notes Backup ends with Warnings due to extra DB_ItemFinishStatus(). Skip extra DB_ItemFinishStatus() call
change: Call i_CHECK_LICENSE_ALL() with string buffer and declare LIC_DATA char szMsg[MESSAGE_LEN]
change: Running sm_backup on CLI must insert result entry and return correct results.cnt
change: LIC_DATA szMsg [MSG_LEN] (34500)
bugfix: Avoid problem with stack size be reducing LIC_DATA messages to MESSAGE_LEN 1536
change: USEFILE: Set return message
bugfix: Avoid problem with stack size be reducing LIC_DATA messages to MESSAGE_LEN 1536
change: USEFILE: Set return message
bugfix: (#18332) - MS SQL 2012 backup is running endless. Implemented the error handling mechanism for ODBC functions
change: enhanced the trace message in the SLU check.
change: removed compiler warnings.
change: slu it is possible to disable encryption if tape is not in drive
bugfix: Add 'vm_server_type' to SQL client select statement
change: slu reports OPEN_ERROR if device cannot be opened and not ENCRYPTION_ERROR
bugfix: (#18278) remove char variable - only wide char left
change: Queue state 'synchronized' instead of 'synchronize'
change: (#18350) SM_CONFIG_DRVIES: RDS DataStore drive queues remain in state synchronized job limit 118350 - SM_CONFIG_DRVIES: RDS DataStore drive queues remain in state synchronized job limit 1. In case of sm_config_drives the special resource 'all' is given to synchronize the queues (requesting all resource queues before configuring the RDS drives).
bugfix: Last commit added wrong arguments for trace message
bugfix: (#18278) Backup of MSSQL database with Umlaut fails, if source "all" is specified
bugfix: (#13971) took out the CHECK_ENCRYPTION section in the CHECK_MEDIUM_FOR_BACKUP since two backups won't work this way.
bugfix (#9135): do not use retries and timeouts when performing tape online check
change: Set UPDATE_TIMEOUT_DATE for new 4.4.2 licenses to satisfy old kit update checker
change: Use GitID in usage output
change: (#13971) inserted more log message for encryption.
change: ARGUMENTS(): Check argument counter revised
NEW: project file for MS SQL 2014
change: Allow command: 'sm_qm -h'
change: sm_qm now supports option '-v {loglevel}' (2)
change: Increase optind always
bugfix: (#18034) VDDK 6 (resolved Linux issues)
change: sm_qm now supports option '-v {loglevel}'
change: remove the additional single quotes which have been inserted for the Uci and insert an empty string if the Uci is empty.
change: remove the additional single quotes which have been inserted for the Uci
change: Do not interpret arguments for submit command
change: Linus: Compile errors resolved
bugfix: (18341) New sm_qm from 06.05.2015 has problems during connect to a pipe. Use WaitNamedPipe() if CreateFile() returned with ERROR_PIPE_BUSY.
bugfix: (#18339) stop further processing if flag i or u is specified
bugfix: (#13971) set the encryption flag and key every time an init is executed and if the medium already exists in the database update those fields. In case a medium is introduced into the sesam db for the first time these variables are passed to the INSERT_INTO_MEDIA method in inserted into the media table by the sm_arch.
NEW: sqlmgr.dll is building for SQL 2014 now
change: do not preserve hash and encryption key when medium is initialized
bugfix: (#13971) update the contains_encrypted_blocks field every time an init is executed and the column in the database is already present.
bugfix: (#13971) set the contains_encrypted_blocks if the medium has been initialized before.
bugfix: (#13971) set the contains_encrypted_bloacks column when executing INIT_VIA_SMS with bFirstInit set.
bugfix: (#13971) use IS_YES makro instead of comparing to the string itself in case there is an inconsistence in the database.
bugfix: (#13971) The encryption is now cleared before an initialization takes place, to ensure that the tape header is always not encrypted. Additionally the encryption is checked if the label is present in the hw_drives table, in case the tape encryption has been cleared.
change: change/adapt code with using iterator of CVssComponentCollection class;
bugfix: implemented the iterator for CVssComponentCollection class; fixed some warnings; changed/adapted code by using iterator;
bugfix: (#18330) display drive encryption capabilities regardless if medium loaded or not
bugfix: (#13971) the encryption_parameters_set was already read from the slu output hence use this variable to check if encryption has been turned on.
bugfix: (#13971) only execute the encryption check if the get label returned online or no trailer.
new: replace the O&O with the newest version 9.292
bugfix: (#13971) reduced the parameters of the CHECK_ENCRYPTION and use this function on several places instead of using SET_ENCRYPTION. Introduced also several checks so methods don't have to be run through several times.
bugfix: (#18218) do not report error if attribute cannot be read
BUGFIX (#18326) Correct command to recoginze all NSS volumes during single item restore over 'vSphere' backups
change: store hash correctly and not just part of it
change: fix initialization of an empty medium
bugfix: (#17923) display some imoportant queue parameters on job start/terminate (allocation, job limit, status)
bugfix: (#18295) sbc_vmware_mount cores, when trying to mount VMDK on Windows
new: (#18325) BSR Windows: O&O DiskImage Backup and Disaster Restore - Generate Network Server settings via sbc_bsr. Add pwd_dll.py a temporary program as a POC for this issue.
new: (#18325) BSR Windows: O&O DiskImage Backup and Disaster Restore - Generate Network Server settings via sbc_bsr. Add oodipwdsec.dll and header.
bugfix: (#9135) - Tape Online Check durch SMS
new: Check output from 'sm_qm show' for '^Name: ' and start processing after this line to ommit any log or debug output
bugfix: C++ header file name has been changed, but dependent file was not corrected.
change: Use oodicmdc.exe
change: (#18321) BSR Windows: O&O Disaster Restore - Restore with Size and Offset - Segment file necessary. Use device node
bugfix: Add space at begin/end of job name if job entry is searched in 'sm_qm sho/name' output to get correct line. TODO: Implement it in 'sm_qm.py' correctly
change: e_GET_MEDIUM() New argument size_t iSize // < size of cpMsg
change: (#18321) BSR Windows: O&O Disaster Restore - Restore with Size and Offset - Segment file necessary. Copy segment {saveset}.dat file to RDS 'gv_rw_smslis:' to allow restore with size and offset
bugfix: (#18307) sm_lis_items application error during lis creation. Avoid error STATUS_STACK_OVERFLOW by increasing STACK size to 4000000
bugfix: (#18307) sm_lis_items application error during lis creation. Avoid error STATUS_STACK_OVERFLOW by increasing STACK size from 100000 to 400000
change: Comment about saveset format now correct
change:(#18102) Mechanism is implemented but it doesn't integrated to SBC.
BUGFIX: (#18316) Binding to STPD data port can fail on Windows, if optional firewall port range is specified. After error WSAEADDRINUSE the recreate tried to connect with IPv6 address. Now allow dualstack communication
change: (#18315) Keep sbc timestamp files in 'gv_rw_times' directory
bugfix: (#16334) for windows ctrl and sshd are started using the new sm_qm option submit_proc_only. This way a remote update is possible.
new: sm_qm submit_proc[_only] allows to avoid stop of whole process group, e.g. for CTRL and SSHD
BUGFIX: (#18312) Return with error if number of source items > 256, instead of silently ignore them. TODO: 'find' can work with much more arguments
bufix: Since commit c2e2833 the PIPE_OPEN_RCMD() did not read stdout stream on Unix. Now after commit 8f2a40b the iPIPE_RAD() must only return -1 in case of 'read from other channel'. In case of error '0' (like EOF) is returned
bufix: Since commit c2e2833 '(#18305) sm_backup did not end due to 'ReadNamedPipe: got error code 6...' the PIPE_OPEN_RCMD() did not read stdout stream if data on stderr was processed before stdout
change: PIPE_CLOSE: Avoid message 'GetExitCodeProcess returned 259 STILL_ACTIVE'
change: Before calling PIPE_... functions the DB_PROC_PIP must be set to {0}
change: x_Semaphore() Add file name to trace message if CreateFile failed
change: Some compiler warnings resolved
change: CtrldAccess() Correct check for option '-d'
new: inserted the compilation of the VDDK 6.0.0.
bugfix: Output of sbc truncated. Revert change commit 5a76e09: change: i_PIPE_OPEN_RCMD(): Add '2>&1' in case of binary mode
change: Use given arguments to build DiskImage xml file
bugfix: (#13971) moved the encryption setting from GET_LABEL to e_GET_LABEL_VIA_SMS if the tape_label could not be read.
bugfix: (#18287) Set '\\.log$' as default exclude for vSphere backup
Revert "bugfix: tapelabel and tapeid may not be null or empty when returned from the sm_sms_interface getlabel."
change: Since commit 3655c39 'change: Windows: Use binary mode for file transfer' the Windows log files have different line ends CRLF and LF. Revert Windows specific 'set global file open mode to O_BINARY'
bugfix: (#18303) VM Restore failed due to corrupted OVF file. Binary file processing added for ctrlc_lib
bugfix: tapelabel and tapeid may not be null or empty when returned from the sm_sms_interface getlabel.
bugfix: (#18303) VM Restore failed due to corrupted OVF file. Now with binary mode and memset(buffer,0,.) the file is transferred correctly
change: iPIPE_OPEN(): Trace messages revised
change: i_PIPE_OPEN_RCMD(): Add '2>&1' in case of binary mode
change: Trace messages revised
change: Usage changed
CHANGE: (#18241) Speed up SQL for getting latest backup state of all tasks. Use triggers on table results to update tasks.results_sts
change: Read 'media.uuid' in case of no explicit label is set for init
change: workaround for bug 18293
bugfix: (#18034) VDDK 6 (resolved missing dependencies)
BUGFIX: (#17672) - Incremental Hyper-V backup. Adapted solution to the changes of using VSS API in case of Hyper-V parallel backup.
change: CtrldAccess() If stderr is requetsed then option '-d' must be set
bugfix: (#18296) stop the SEP sesam BSR Pro before the uninstall of the SEP sesam BSR Pro happens. Start the service after the update has been finished.
bugfix: (#18296) SEP sesam BSR Pro is now stopped, started and killed using sm_os_startup
bugfix: (#18305) sm_backup did not end due to 'ReadNamedPipe: got error code 6 (ERROR_INVALID_HANDLE)'. PIPE_OPEN() now checks this case. PIPE_OPEN_RCMD() end loop if (iRetStdout > 0 || iRetStderr > 0)
bugfix: (#13971) If tape trailer could not be read but tape header could be read due to encryption than do 1 retry of reading the tape trailer. If it fails execute with tape trailer corrupt.
bugfix: (#18290) Reinsert the PACKAGENAMECHANGED and all other needed variables so the uninstall and install only takes place if the package name has changed.
bugfix: (#13971) in case an encrypted tape is reinserted the tape trailer read fails and the tape header has to be read. This will not lead to an error for encrypted tapes.
change: Generate BSR xml file with BOM (utf-8-sig)
change: removed the PACKAGENAMECHANGED variable for the Uninstall and changed the option value to 18. Possible uninstall install fix.
sbc_jira: log
change: Use new BSR PRO comand 'oodicmd.exe /script {file}'
change: (#18303) VM Restore failed due to corrupted OVF file. Set binary mode explicite in REMOTE_COPY_CLIENT()
change: Add 'B' to getopt string
change: Moved the relcalculation of the readcheck state before setting the status to get the execution state in the end of the trace file
new: Call 'strip()' for all lines read from 'sm_dedup_interface' call
bugfix: Search VMDK data file using wildcards, because 'sm_vmdk_interface replace '@' sign with '_', but path of saveset file isn't changed
bugfix: Kill the java*sm_ui.jar, javaw*sm_ui.jar, sm_java and sm_javaw process when sm_os_startup KILL is called
bugfix: (#18299) BSR (DiskImage) not installed: Missing registry tree. Use self.bsrPath to set pathes for BSR CBMR
bugfix: after keys regeneration backup over ssl does not work
bugfix: (#18299) BSR (DiskImage) not installed: Missing registry tree. If BSR DiskImage is missing then go on with check for BSR CBMR
change: inserted the HKLM key before the registry entry.
change: use the new O&O version 9.291 instead of the old one.
change: always update OODI no matter if the version changed or not just for testing.
bugfix: (#18290) inserted the set of the registry options for the RDSH before the key is deleted otherwise the key cannot be deleted.
change: Windows: Use binary mode for file transfer
change: resolve_path() with trace
change: i_GetPrivateProfileString() use REMOVE_LF() for retrieved value for given key
bugfix: (#13971) insert the hash key base 64 of the encryption key in the media table during init. Enlarge the hash key base 64 variable to a size of 180 since it is larger than the binary hash key (128)
bugfix: change %d to %s for the string variables.
bugfix: removed obsolete code which set the encryption during the GET_PREPARED_MEDIA.
change: another try to get the installation of the O&O working.
change: Use 'sm_popen' to call '.cmd' on Windows. (sm_Popen could not start shell scripts)
bugfix: Set variables to NULL in 'getlabel' function, so they are NULL in case of 'pscanf' failes
change: inserted the error message if the deletion of the registry key fails.
change: (#17807) BSR Windows. The 3rd BSAGetData() call failed with 'third NO_MORE_DATA >> return BSA_ABORT_SYSTEM_ERROR'. Reset the iAlreadyNoMoreData counter if valid data was read
NEW: (#18121) Zarafa: Do not exit with an uncaught exception, if parent rescource couln't be found during creation of mail LIS file
bugfix: remove the registry key for the remote desktop service host that was inserted during the install/update after install/update has been finished.
NEW: (#18133) Use 'sm_Popen' to get correct return code on Windows during copy VMDK files
bugfix: convert the sha-512 hash key to base64 before passing it to the sm_sms_interface.
change: Do not change process id/group to Dovecot user in case of backup
new: Write new UUID on tape during init for media without UUID in header set
BUGFIX: Set exit code > 0 if an error has occeured during 'sbc' execution
change: remove the check for the CLEAR_ENCRYPTION output which is not needed anymore. This is a left over.
change: (#17807) BSR Windows: O&O DiskImage Backup and Disaster Restore. Use parameter <ignore_invalid_free_size_appointment>yes<...> to avoid size checking.
change: BSR_Windows: INQUIRE savesets. If saveset is given then return only information about this saveset.
change: sms - extended tape header - remove whitespace on SMS version
change: new major version 4.5A1 due to format change.
change: Link sm_arch: SM_RECOVER_LIB is already part of SM_COMMON
bugfix: (#18290) save the variable for the rdsh_registry in the execute_sequence ini file whose values are read in the UI sequence.
bugfix: For saveset recover on data stores: Remove key 'sesam_server' from DB 'INSERT' statement, because there is no such column in 'DB:results' table
BUGFIX: (#18232) sm_arch: Media Recovery does not create external queue job for watch process. Submit sm_sbc_com_ext into drive queue
change: use the newest version of O&O 9.290
bugfix: (#18290) disable the RDSH (Remote Desktop Session Host) during install/update/uninstall since this there is an incompatibility with the Embedded MSI technology and the Windows Installer Coordinator.
* bugfix: Hyper-V backup on standalone host with message: the sm_sms_backup ended with Status= X  Msg= Subprocess sm_sms_backup undefined broken
NEW: 'Saveset tree': Support multiple cloned savesets
change: License output for 'MSSharePoint GB ( Microsoft SharePoint y*500 + Microsoft SharePoint Extension GB x )' added
change: (#18282) License output must show the sizes for Volume Front Side and Storage. Show FrontSide and Storage always in license output
change: (#18275) Windows: Kernel module SQL access to SQLite database returned with error 'database is locked'. PROT of DB commands only if debug.ini PROT_DB > 0 is set
change: sm_sms_interface replace -o reinit wihh -o over
change: Added sleep(seconds) function, example: 'SELECT sleep(10);'
change: Revised statement to fill user_group_relations with (user_id,group_id) = (1,1)
bugfix: (#18285) Avoid sm_sms_backup core dump in case of HyperV backup ends with 'Subprocess sm_sms_backup undefined broken' due to browsing node '.'. Extra logic added to abort backup in this case
change: do not et encryption flag to 1 if -E 0 given
change: extended tape header, store SMS version in short form and replace commas in Sesam version
new: Return more information like FDI type during query of vSphere savesets
change: sms extended header - handling for UUID on old media when using -o reinit
bugfix: use the correct parameter type for the function pscanf. char* instead of char[]
change: Use M_STRLCPY()
change: stpd_sms_stor_i() Use 'strlcpy(szError, sBackup.ReplyMsg, STPD_REPLY_LEN)' to avoid string truncation
change: Avoid to update hw_drives if nothing changed
change: Use cm_trace.h macros LOG_LEVEL, LOG_VERBOSE, SET_LOG_LEVEL(A) and SET_LOG_VERBOSE(A)
change: x_Snprintf() Avoid extra trace
new: Call 'sm_sms_interface init' with option '-o reinit' if not called for introduce media
new: Allow replication of one saveset to multiple destinations
bugfix: removed the Shell parameter from the O&O installation.
change: OO_VERSION=9_0_289
bugfix: compiler error removed.
bugfix: (#13971) only insert the uuid, encryption flag and encryption hash when first initializing a medium.
correct message number
bugfix: (#18280) Remove 'PROT' statement with wrong number of arguments
Merge remote-tracking branch 'origin/master'
bugfix: (#13971) Do not clear the encryption after backup. The encryption is cleared when the medium is unloaded or when the device is plugged off. Decrypt the given key and encrypt it differently before passing it to the slu. Use the same decryption for the slu which is used when calling the slu from sm_get_volume_lib
change: set disposition before reading label when -o reinit option is used
change: use same timestamp for creation date and last init date
* bugfix: Hyper-V browse operation is failed on the standalone server with message:  sbc-1501: Error:   CVssCluster::CreateWmiObjects: Could not connect to server. Error code = 0x80041008
bugfix: Set 'encryption hash' to empty stream if no encryption is set
change: BSR_Windows: INQUIRE savesets: Count only savesets with data_size>0
change: sms added attribute first initialization
change: removed cause for warning C4214: nonstandard extension used : bit-field types other than int
change: BSR_Windows: If INQUIRE for single saveset then return saveset information together with size
BUGFIX: (#17822) Restore failed due to wrong tape block size. The device supports maximal block size 16MB but SMS maximal buffer is 1MB and even with tape block size set to 64k the ReadFile reads several blocks at once. Now set read buffer size to 64K (=uiMaxBlockSize)
change: Compiler warning solved
change: QUE_SUBMIT: snprinf() on Linux does not allow to use target string as source
change: Wrong argument for s_Snprintf()
change: GET_TAPE_SERVER() revised
change: Added function void trace_ln( char *cpLine )
change: Code revised in sm_lib.c: string operations and unreferenced formal arguments
change: x_Dump uses trace_ln instead of WriteLog
change: encrypted flag set to false by default
bugfix: fixed SSL certificate verification
change: Remove spaces from key words in 'sm_sms_interface' output
change: (#18275) Windows: Kernel module SQL access to SQLite database returned with error 'database is locked'. Added PROT messages to identify locking SQL statement
bugfix: Correct SQL statement to select a single saveset for replication
bugfix: (#13971) insert the column last_label for the last loaded label.
new: (#18049) vSphere 6(supported sslThumbprint for linux)
bugfix: (#17752) sm_qm_main log becomes huge the more drives are configured. Avoid log messages in function WalkRandomExecQueues()
change: iSEARCH_SAVESETS_BY_SOURCE_TASK() If job is set to '*' and inititator is 'BSR_Windows/Linux' then inquire all tasks for given initiator
bugfix: (#13971) pass the media pool to the drive_handler when calling the drive_handler with option init to obtain all encryption flags correctly.
new: (#18049) vSphere 6(supported sslThumbprint for windows)
change: sm_sms_interface no base64 check if E = 0
bugfix: (#13971) Changed the parsing of the tapelabel, tapeid, uuid, encrypted flag to regular expression. This is a lot easier to understand.
change: Compile error resolved
change: sm_sms_interface getlabel; remove space between Tapelabel and TapeID
bugfix: erase two lines which were accidently left.
CHANGE: increase the SMS_VERSION and the SMS_VERSION_RC to 4.4A4 and 4,4,1,4 since the sm_sms_interface output contains more info than before.
bugfix: (#18238) 'CoUninitialize' doesn't hung anymore.
bugfix: put the hyper-V specific code in a if def block so other backups are not affected.
bugfix: try to fix parallel hyper-V backup. The method CheckSelectedWritersStatus was missing in the Hyper-V specific parallel execution.
* change: (17272) Merge stpd_null_stor and stpd_fs_stor
bugfix: (#13971) pass the cryptographic hash of the password used to encrypt the tape to the sm_sms_interface init. Use SHA512 for creating the hash.
bugfix: execute a initialization of the medium if the flag over is set no matter if the label is the same or not.
change: Trace message added if ReadFile returns buffer with size > 64K
BUGFIX: (#17822) Restore failed due to wrong tape block size. The device supports maximal block size 16MB but SMS maximal buffer is 1MB and ReadFile() returns up to 16 * 64K blocks within 1 read operation. Now set fixed block size to 64K.
bugfix: (#18213) insert filter for cluster or non-clustered VMs (fixed backup ALL case)
 change: Call $(MAKE) -f makefile.win32 check_version in $(CM_COMMON_DIR)
bugfix: (#18204) Perform safe search of TT by seeking to EOD and sequentially read back; handle EOM area correctly by catching ENOSPC -- update to work on master
bugfix: insert the build of the version_h to the prepare_version routine.
change: insert the define of the _HYPER_V_BACKUP_COMPLETE_AT_EARLY_STAGE variable just to test the changed order.
bugfix: (#13971) use the new DRIVE_HANDLER_PROP structure to pass the uuid and the encrypted flag from the drive_handler back to the sm_arch after the tape has been initialized. This information is than written to the db.
change: (#17506) added '-o reinit' option, added support for CryptHash-base64, added Sesam version and SMS/STPD version
change: introduced a new structure for the drive_handler called DRIVE_HANDLER_PROP which contains all necessary information for the function drive_handler.
change: (#17807) BSR Windows: O&O DiskImage Backup and Disaster Restore. Call oodiag with option /run {script_file}
change: Add git commit SHA to bsainit and bsaqueryapiversion
bugfix: (#18223) ME10 Restore to generated RDB - Error - Exchange Server warning: CExchangeServer::Restore - MslFileParser::Parse(): Failed to match string from msl file. Since fix #18161 'Selective restore retrieves only first part of selected volume if data is separated by other volume' the msl file and blow up for exchange was corrupt
bugfix: try to make the install condition unvalid in the first chained installer package.
change: added the variable OODIINSTALLED in the log message.
bugfix: (#13971) during media initialization write the uuid and the encrypted flag in the tape header using sm_sms_interface. According to these values obtained by sm_sms_interface getlabel encryption is switched on and off when reading and writing on tape.
bugfix: (#13971) create and insert a UUID in the media table this UUID is than written in the tape header via the sm_sms_interface. Move the part of the encryption specific columns from the media table to the section of the new creation of the media row since these values can only be set during media initialization.
new: inserted the UUID in the MEDIA_PROP structure.
change: sms extended tape header refactoring
new: Add VIEW 'tasks_clients' cotaining all 'task' joined with 'clients' and 'locations'
CHANGE: Calculate balance for dedup enabled volumes by comparing 'high_water_mark' and 'used' (disk space according stat).
bugfix: ignore single items in case when entire user mailbox is selected to be restored.
change: Allow up to 255 columns for 'select'
better trace
bugfix: (#18238) Wrong VSS-backup methods wrong call order has been excluded (with easy swithcing between correct and wrong order)
change: print tape parameters correctly
NEW: Call 'sm_config_client' for Sesam server itself, to update 'DB:clients' with version number
BUGFIX: (#18252) Set SDS service to off if no Si3 dedup store is configured on a RDS
change: added log message about vss reparse point on default log level
accept 'SELECT' as token too
new: Remove double '/' from data store path
change: (#18166) BSR Windows: O&O Disaster Restore. Return saveset info of successful backups with linefeed instead of ';' delimiter
change: iDB(): return message from called iDB_ACCESS() e.g. result of aggregate function
bugfix: (#13971) pass the encryption value to the sm_sms_interface init if the encryption flag is on. Also write the UUID in the tape header and the database.
new: (#17807) BSR Windows: O&O DiskImage Backup and Disaster Restore - Call DiskImage via sbc_bsr. Create job xml file for BSR Pro. TODO: Start oodiag.exe /RUN {job_id}
change: (#17506) following attributes implemented:  UUID attribute read/write, encryption attribute read/write, usage counter read/write, SMS version read/write, STPD version read/write, Saveset info prepared
new: (#18213) insert filter for cluster or non-clustered VMs
new: Set 'DB:tasks.results_sts' to 'NULL' if no results entry for a task exists
bugfix: (#13971) drive_group is set encryption_capable if all drives are encryption_capable. Drive is set encryption_capable if slu ... -E get returns that the drive is encryption capable.
add pre remove script for Debian Sesam GUI
bugfix: (#18223) - ME10 Restore to generated RDB - Error - Exchange Server warning: CExchangeServer::Restore - MslFileParser::Parse(): Failed to match string from msl file. Adapted client according to new format of the msl file
PostgreSQL 9.4 for Debian 8
BUGFIX: Check, if 'sm.ini' really exists before try to 'source' it. During GUI installtion on Linux '/etc/sesam2000.ini' is created 1st, but 'sm_set_java' is called before 'sm.ini' creation
bugfix: skip directories and system files in descriptors scanning
bugfix: (#18231) When reparse point points to other volume then the whole volume is backed up
CHANGE: Remove SMS directory from Sesam default 'LD_LIBRARY_PATH'. Shared libs in SMS dir are searched by 'rpath' in the executables
change: (#18241) Speed up SQL for getting latest backup state of all tasks. Execute trigger only if results.state was changed
change: (#18241) Speed up SQL for getting latest backup state of all tasks. Trigger for psql adapted
CHANGE: (#18241) Speed up SQL for getting latest backup state of all tasks. With trigger the explicite 'update tasks.results_sts...' are obsolete.
wip: Try to find 'total' lines wfor SQL statements
new: Return exit code '0' if 'sm_sho dir' is called for a client with access mode 'PROXY' and client is a 'vCenter' or 'ESX Server'
CHANGE: (#18241) Speed up SQL for getting latest backup state of all tasks. Use triggers on table results to update tasks.results_sts
change: Since de57d140 'extended buffers for xbsa internal use' this implementation was not longer conform to XBSA Standard C425. Now the API Version Issue is set to 2 to indicate the difference to 1 (C425 conform).
bugfix: inserted values for the order of the chained packages. First uninstall and second install the O&O package.
remove obsolete trace
new: Trim VADP config string
change: Get diskImage installation path via registry
bugfix: always deinstall O&O and reinstall it in case the user chose this option.
change: iDB(): Call DB_SEL_LOOP_RECORDS() for local structure DB_REC_LIS
change: (#18166) BSR Windows: O&O Disaster Restore. Return saveset counter of successful backups without old BSR (CBMR) savesets (source ='BSR.000...')
change: sbc_com_interface: If job name is not given then set job to '*' to query all tasks for given backup type (initiator)
bugfix: use the BSR_PRO instead of the BSRPRO variable and remove the default value of the variable otherwise the property value is not passed to the execute sequence.
bugfix: (#18231) When reparse point points to other volume then the whole volume is backed up (correct relative raprse point path)
change: Use 64k default blocksize for writing to 'NULL' in STPD
change: Add prototypes
change: Use com_log( 3154, ... ) to log assembled STOR/RETR commands
change: trace (char *cpFormat, ...) with log level prefix '%l# ' support
change: Function com_util_u.c _trace(...) removed use trace(...) instead
change: Usage with option -D (delete saveset)
change: remove the default value of the OODIINSTALLED property.
change: write also stderr to sm_db_update_update_sbc log file.
bugfix:(#18231) When reparse point points to other volume then the whole volume is backed up(root directory is reparse point)
change: (#18166) BSR Windows: O&O Disaster Restore - Retrieve Saveset Size. Inquire now return size={bytes} for BSR-Windows and allows to inquire clients and tasks if task is set to '*' or ''
change: Use ' s..-' to position to message code in NOT file
new: (#18230) Remove old 'glbv'  written by restore and get volume
change: Set SEP_DeDup_REP for edition DEMO, NFR and Volume
Non-zero return value from cmd_fsck_status when fsck aborted.
bugfix: (#17856) Avoid error message during "ValidatePlannedSystem" call in case of restore Hyper-V VM with corrupted snapshot.
new: Add '*results.mtime' to VIEW 'all_results'
NEW: Check for abort message from Si3 'fsck' and report this as error
Revert "bugfix: (#17856) In case the asynchronous job is called using ValidatePlannedSystem and this job fails a warning instead of an error is returned."
new: Add mtime for all 'results' table
new: (#18167) BSR Windows: O&O Disaster Restore - Retrieve last 200 MB of the DiskImage backup saveset. SBC allows to restore single item with offset and size: -n <segment>[:<offset>:<size>]
change: Messages added to trace the single selection (offset+size)
bugfix: (#18221) sm_qm_main cores on newday event. Use Critical Section object to avoid concurrent writing of StopProcessGroup() threads into sm_qm_main log file.
bugfix: (#17856) In case the asynchronous job is called using ValidatePlannedSystem and this job fails a warning instead of an error is returned.
bugfix: (#15652) Hyper-V backup concurrent processes synchronization
change: insert the correct copy statement in the echo command.
bugfix: (#18221) sm_qm_main cores on newday event. Check if pipe handle is valid. Use argument pPeer->iStatusFlag in ANSWER_WILDCARD()
NEW: insert user name and password parameter to the parameter list of the install sequence for windows. The user and password are inserted in the database if they are given and will be reused during update and other operations.
bugfix: (#18203) VSS backup returns path to original location and data is not read from shadow copy if directory contains a reparse point
change: XX_R_NO_ROW_FOR_TASK with tag E015-RESTORE
Use 24 hour time format for mtime
change: remove the checkout rule from the makefile.win32 since this rule is never used. The correct rule is located in the build directory from the CVS.
change: combine the sesam_db.sql and the sesam_sqlite.sql to one file and copy it to the kit.
bugfix: in case an error occures during slu -s, -E, or -R so not exit but rather continue with the commands and just write in the log file that an error happened.
bugfix: only use COPY for Hyper-V backups since Full, diff, and inc are not supported yet.
bugfix: replace the ; with a , in the sed replacement command.
fix syntax error in 'create results'
change: Adapted javasqlite4.mak --> makefile.win32
new: Upgrade to javasqlite-20140624
change: renamed the function i_READ_SLU_S_OUTPUT and struct SLU_S to i_READ_SLU_OUTPUT and SLU since they contain information about several slu calls not only slu -s
change: replace the current_timestamp during the build process for sqlite with datetime
bugfix: (#18201) - MSSQL backup crashes. Corrected the call of FreeSplitList() in according to recent changes.
new: Set default value for 'DB:results.mtime' (creation time)
correct name of trigger
bugfix: put the two or conditions in brackets since the and condition is the most important one.
change: Call FreeSplitList(&pszItems);
change: Avoid trace with 'WARNING: unknown license key ..' for unknown license keys. Only if the value is unequal to 0000 a warning about not supported license key will be displayed
change comment
new: DB specific Sesam DB SQL files (PGSQL/SQLITE)
new: 'DB:results.mtime' to record modification time of records
change: insert description for the device_encryption_capable and tape_encryption_capable values.
change: Set Java debug listen port to '5006' to be not the same as used for server (5005)
BUGFIX: (#17718) Now if JAVA_HOME is not a valid version the registry is checked for windows to obtain a valid java version.
new: Restore '-n' option now with optional saveset offset: -n <segment>[:<offset>]
change: (#16989) Git commit must be visible. Make rule to create sesam_version.c
change: (#16989) Git commit must be visible. 'char version_string[]' already defined in version.c
change: (#16989) Git commit must be visible. The version string is now build from SMS_VERSION and git commit SHA. The SMS/STPD log files will show complete git info in extra line.
NEW: (#18199) Set EOL of automatically created restore tasks to 2 days, so they will be deleted after this time
Fix message text
Log message concerning import VM without snapshots has been changed + some minor changes
new: const char gszGitCommit[] to keep recent git commit SHA
BUGFIX: (#18198) AIX: if tape trailer fails to write, sm_sms_watch hangs in CHANGING_MEDIA state. Set GLBV gv_sms_eom_{device} to 'BREAK' for older STPD program
bugfix: (#18190) Dovecot mail restore create SEL file which leads to a STPD core. Use strlcpy to avoid string overflow in STPD.
BUGFIX: (#17957) If StartSnapshotSet or AddToSnapshotSet exited with VSS_E_SNAPSHOT_SET_IN_PROGRESS the retries did not work as expected, since the break condition exited the loop in the first case. Now the loop is executed until the break condition is reached.
sbc_jira: error handling
change: (#16989) Git commit must be visible. Display git id as 'Build: ...' for VSS DLL. Example: sbc-3036: Info:  # DATABASE SUPPORT: Volume Shadow Copy, VERSION: 4.4A3 Build: fda7e08, Released: Apr 13 2015 #
sbc_jira: trace
change: Trace log level 2 for 'message code is not in sbc format: ...'
sbc_jira: error handling
BUGFIX: (#18176) Dovecot mail restore create SEL file which leads to a STPD core. Changes in sm_restore commit 7b96b3b for call #18161 skipped 2nd BLOW_UP_SLL() function call
sbc_jira: testing, branch
bugfix: changes in the trace logging
Hyper-V restoration from corrupted saveset
bugfix: revert code related to functionality of Exchange reformatting lis (60deebfd24b70c039312b8c39803a8ceda8051dd); united Exchange functionality - reformatting and validation into one module =- sm_exchange.py; revert code on the client side - vss module
new: (#18019) New License Key MSSharePoint_Extension_GB. Calculate MSSharePoint GB = MSSharePoint_Extension_GB+MSSharePointServer*500 GB
change: Correct offset +3 to get sbc message code
bugfix: (#18179) Server verification failed during backup over https
change: Usage for dismount option 'label' to set label in hw_drives to NULL
change: Trace with milliseconds only if define TRACE_MILLISECONDS is set
change: ANALYSE_BACKUP_LOG/RESTORE_LOG() now accepts date time with milliseconds too
BUGFIX: (#18184) sm_sms_watch: media dismount on remote device servers on AIX will cause other backups to fail if shared drives are used. Getting tape status on AIX will occupy the tape device and therefore other shared drive operations will fail. Avoid inquiring the tape status - do only remove the memorized label in hw_drives.
new: Trace lines with milliseconds
change: changed the BSR_PRO variable name to BSRPRO.
BUGFIX: (#18181) WORM medium backups are deleted from DB in case EOL expires. Skip delete LIS and saveset entries in table results/_lbls in case of WORM media
change: (#18117) Maintenance license key. If MAINTENANCE_TIMEOUT_DATE is set to 0 then use UPDATE_TIMEOUT_DATE instead
bugfix:(#18107) added ability to use port range for FTP and HTTP protocol for Windows client.
source code format
bugfix: (#18174) Return dummy falues in case of error to get the expected number from caller
bugfix: (#18151) Inserted a checkbox for the SEP sesam BSR Pro installation. Hence if the user deselects this option SEP sesam BSR Pro will not be installed or deinstalled if it has already been installed.
change: (#17384) License for loaders with more then 9999 slots. Special check for SLOTS licence old unlimited value 9999 and new unlimited 99999
* Bug 16206 - Validating Backup Integrity By Using CHKSGFILES - take off the mechanism of catching errors of Exchange validation function
change: temporary roll back bugfix 18107 to resolve problem with FTP
new: (#18020) New License Key SEP_DeDup_REP. Use calculated value in license check
NEW: (#18020) New License Key SEP_DeDup_REP. Calculate the count of Replciation servers based on replication tasks
re-activate Exchange log file validation
* Bug 16206 - Validating Backup Integrity By Using CHKSGFILES - implemented and used the join_path() function instead of os.path.join() because of issues of using on the UNIX system
bugfix: in rear cases when MBR signature starts from zero script doesn't work. Adding "0" in front until length equal 8
bugfix:(#18107) added ability to use port range for FTP and HTTP protocol for Windows client.
change: use the correct variable for the registry Name of the O&O registry entry.
new: (#18117) New SERVICE_MODALITY 'Maintenance'|'Subscription'
Bug 16206 - Validating Backup Integrity By Using CHKSGFILES - added even more logs
change: sqluvint() trace gszGitId added
bugfix: Link error 'cannot resolve cm_split' resolved
change: Trace added for saveset offset
change: use the correct registry field type for the o&o registry key.
change: In case of relic of an exiting file set target version to old version if not an upgrade or downgrade was requested
change: Compiler warnings resolved
new: (#18117) New key MAINTENANCE_TIMEOUT_DATE only if target version is v4.4.2
change: Link with $(CM_COMMON_LIB) to get cm_split() function
change: mbstowcsl(): Use cm_split(cpLocales, ",", &ppszLocale) to allow to work with list of locales. An empty entry stands for the current locale. So ".UTF-8," will first try to use UTF-8 and then uses current locale to convert MBS to wide char
bugfix: VSS API sequence slight modification
remove call of 'disconnectdb', because it leads to a core in case of a SQL syntax error. This has been introduced to avoid memory leak if PSQL server isn't up. TODO: How to handle both problems correctly
change: add some log messages.
bugfix: Correct SQL statement to set toal data size for group result
change: add some log messages.
bugfix: do not use a backslash at the end when copying a file.
Client build fixed for Mac OS X
bugfix: linux packdata build
change: use the correct registry key to detect if O&O has already been installed.
In 'get_pool_from_label' remove memset for langth of pointer. Memset to '0' is done in calling function
bugfix:(#18041) fixed warning on backup (both COMPLETEBACKUP and POSTBACKUP phases) types which needs ASR Writer to get data doe diskpart script.VSS_E_WRITERERROR_INCONSISTENTSNAPSHOT
change: insert the copy of the cm_crypt.dll in the makefile.win32 of the sbc_w32_os2\sbc and the sbc_esx\sbc_vmware directory.
change: insert rules for copying the cm_crypt.dll to the sbc and kernel directory.
change: copy cm_crypt.dll also in the sbc directory for sbc patches
bugfix:(#18041) fixed warning on backup types which needs ASR Writer to get data doe diskpart script.VSS_E_WRITERERROR_INCONSISTENTSNAPSHOT
bugfix: shrunk warning during sbc_vmware SAN backup
bugfix: (#18138) - ME10 - Backup - error - sm_reformat_lis returned exit state 1. The logical error is occurring in case of selective backup of one or few dbs.
NEW: Calculate 'balance' for data stores on deduplication volumes using free disk space only
BUGFIX: (#18161) Selective restore retrieves only first part of selected volume if data is separated by other volume. Process blow up of selected items on sorted LST file and resort to LIS file afterwards
bugfix:(#17961) Memory leak in DB access. In case of an error 'disconnectdb()' will be called now
preset 'szBuf' to empty string because it's used in a compare operation later
bugfix: (#18160) - Add logging for Exchange backup validation
new: Call 'sm_reformat_lis -s <saveset>' so saveset information can be used for DB access in sm_reformat_lis
new: add option '-s saveset' so easier DB access
bugfix: Remove Exchange validation until it works correctly
new: (#16734) Write of SMS Tape trailer as 64k block. Skip setting block to TAPE_TRL_SIZE_2K in function read_tape_trailer()
new: (#16734) Write of SMS Tape trailer as 64k block. Skip setting block to TAPE_TRL_SIZE_2K in function resume()
bugfix: (#18138) - ME10 - Backup - error - sm_reformat_lis returned exit state 1. Changed SQL query for getting FDI saveset chain.
NEW: (#16460) Call 'sm_alarm' with NOT log a 3rd parameter. Add a parameter to 'i_RCMD_LL' function to switch log into Sesam main log
NEW: (#16734) Write of SMS Tape trailer as 64k block. Write Tape Trailer block with 64K for Windows and Unix
change: insert the new O&O version in the kit.
bugfix: insert a missing semi colon.
new: insert the sm_parse_guilog.exe in the msi package.
BUGFIX: (#17483) Do not execute set_vddk on a GUI installation without client.
NEW: Update 'DB:results.total_size' for running backups
new: If no log file name is set, then parse 'gv_rw_lgc:sm_gui_server.log'
BUGFIX: (#18072) Incorrect syntax in Notes LIS files shows no items in restore wizard. Remove leading '/' before storing EA/ACL of file into cpio archive
jira: disable module in sbc_proxy
bugfix: use the correct registry key to check if O&O has already been installed.
change: do not install the system tray of O&O during silent installation.
new: insert the sm_parse_guilog.exe to the build process for windows.
bugfix: removed some compiler warnings.
bugfix: insert the -c parameter to the browsing of the Hyper-V computers. This should make an all backup possible for the Hyper-V core computers.
NEW: Write complete data size of a multi saveset backup (vSphere VMDK for example) into 'DB:results.total_size' of the main task
bugfix: (#17752) reorganize logs; do not print QUERY type requests
bugfix: build issue linux sbc_esx
bugfix: fixed vmware build on linux
change: Since commit a57d32c '(#17320) MSSQL Backup of database with special characters fails' the arguments in sbc_msg() and trace() are decoded with UTF-8 and not with current locale. Switch back to decode with current locale unless UTF_8_ENCODED is defined
add usage
Use 'getops'
wip: Add 'sm_parse_guilog' to makefile. Sort targets alphapetically
new: Program to parse SQL statements from new Sesam UI server log
change: Set SignalHandler() for DEBUG with x_PrintStack() only if DEBUG and PRINT_STACK are defined
bugfix: (#18141) sbc cores when called with verbose level -v 6. Avoid use of mbstowcsl() - use MultiByteToWideChar() or mbstowcs() instead.
Print out command line including quotes
bugfix: (#18115) If restore fails do not import VM to HyperV Manager
Correct 'requires' for SuSE
change: GPT partition relocation for Windows PE will be done by disk letter instead GUIDs if possible. Extended and fixed log messages.
change: Compiler warning about memset in x_DumpEx() resolved
bugfix: Preset 'gc_msg' and 'fsck_msg' to empty string so no garbage will be inserted into DB:data_stores in case dedup server is down
bugfix: (#18063) use parameter -k to pass the user and password to the sm_remote_installer.
rhevm poc: skip vm with direct lun attached
new: Use RedHat version to set dependencies
skip lun
change: Argument in FreeSplitList now ( char *** ) to allow to set cahr ** pointer = NULL
change: x_Free(void *mem) now sets mem = NULL
NEW: Add 'lsb-release' to RPM dependencies for SuSE
change: Function cm_strtok() moved from common/cm_kerenel.c to common/common/cm_split.c
change: FreeSplitList(..) revised
change: FreeSplitList(..) added
change: insert the SESAM_VERSION in the link rule of the sm_dedup_interface.exe
change: insert the creation of the BINDIR for the cm_crypt.dll copy command.
bugfix: (#18100) Browsing of Hyper-V cluster (ignore domain when comparing hosts)
change: concatenate the alive_test results of all clients present in a hyper-V cluster.
bugfix: replaced all BIGINT fields with numbers before there were strings.
change: Use macro CFLAGS set in makefile.wcommon
bugfix: (#18100) insert the -c clientname parameter to the sm_client dir call. This is needed for browsing Hyper-V cluster and nodes correctly.
bugfix: Move creation of 'users' table before 'locations' which refers to 'users'
CHANGE: Add '/' to path for 'du' command so it will return correct value if path is a sym-link or mount point
change: added the cm_crypt.dll to the gui and client package.
bugfix: added the missing _function_name variable
bugfix: removed the unneeded variables from the run_remote function and inserted the missing variable to the execute_remote_cmd function call.
bugfix (#18138) - ME10 - Backup - error - sm_reformat_lis returned exit state 1. Use 'rb' flag for opening metadata file. Additional fix to 7466b9124c1af543910685d7a772d6474f58e710.
NEW: Set reference for 'clients.datamover' and 'clients.vmhost' to 'clients.name'
bugfix: Read RPEX client over drive node, if not set on command line
new: inserted the user and group specific database fields to the locations and the clients table. Cleaned up the spacing
new: insert the insert statements of the users, user_group_relations, permissions, groups, group_permission_relations in the install and update sequence
bugfix: (#18138) - ME10 - Backup - error - sm_reformat_lis returned exit state 1. Binary read the meta-data instead of read as text.
bugfix: reload rmi only for UNIX. Until now the rmi was off in case of Windows during update hence the reload of the rmi did not take place. To ensure that this will not affect windows update the reload will only be executed for UNIX.
change: Avoid xfree() call
Remove 'RPEX' desktop link during unmount of saveset
change: Read 'Public' folder for RPEX link over registry instead of environment variable
bugfix: start only qm not all services before starting the sm_db_update during update procedure.
change: added -c computername option for sm_client(Example: sm_client dir -c hv-node1 /Hyper-V:)
BUGFIX: (#17916) implemented EOD detection
BUGFIX: (#18131) sbc returns exit code 0 after errors during restore
change: removed the first declaration of the CURL_LIB
bugfix: since the common objects were not linked correctly the free of the url_encode function did not work as expected.
change: Comment for last call of iPackItem(NULL, NULL,NULL,...) changed
change: (#18132) Backup of CIFS network share is to slow especially when backing up small files. Avoid malloc() in iPackItem()
Merge remote-tracking branch 'origin/master' into sbc_performance
bugfix: inserted an if clause for windows for the backup using vss and all_writers.
bugfix: (#18088) If the current directory contains the GUI jar file than this directory is used as GUI directory from now on.
new: added the newly created cm_crypt.dll to the msi package.
bugfix: (#15345) now if sm_qm_main/sm_main start qm is called only the qm is started. Hence the turn off of all services during update is not necessary anymore. Remove the turning off of the services in the sm_setup.
CHANGE: Increase sbc performance with STAN buffer size to 1MB
change: inserted the CM_CRYPT_LIB in the DEPENDS_BUILD for the non static build.
change: add the cm_encode to the cm_common.lib
bugfix: inserted -o vss and -x all_writers as well as the creation of a log file for the backup of the sesam db which takes place right before the sesam db update.
change: changed the field for the server type from Hyper-V to Hyper-V-Cluster
Merge remote-tracking branch 'origin/master' into cm_crypt
change: move the CM_LIB_ALL after the CM_CRYPT
change: (#18100) Browsing of Hyper-V cluster(Added -c computername argument)
change: remove the cm_encode from the build rule since it is already included in the cm_comon
new: Start 'sm_sms_watch' at once after drive configuration
RPEX: Read Sesam bin dir from RPEX client. Use '%PUBLIC%' env variable to get path for desktop link
jira poc: import changes
change: insert the cm_encode.o in the link rule
change: insert the cm_encode.o in the common object file list
new: inserted a new library for the crypt functions. Adjusted all windows build scripts
bugfix: (#18100) Browsing of Hyper-V cluster(fixed stand-alone Hyper-V issue)
bugfix: (#18063) the sm_remote_installer has an additional switch -a which is responsible for the service user. If this switch is given the service is started using the user which was used for the update/installation.
jira: remove jira_common.py
NEW: Jira module POC
call 'shortcut' on RPEX client
copy RPEX cmd file to client
new: 'DB:results.total_size' to store toal backup size for task using sub tasks. (like vSphere: Main 'saveset' includes data size of config, but not from VMDKs)
change: (#17783) increased the size of the message_id field from 64 to 255 since it is possible to insert 251 char in the RSS title.
new: Added relocation using information from VDS for GPT partition restore from Windows PE
strip quotation marks from parameters
change: compile the chksgfiles.exe in case when the appropriate flag is set CHKSGFILES=1
Use complete Exchange DB name for searching
BUGFIX: Archive adjustment: Restrict slot range to slots configured for loader
bugfix: added vssbackupdecorator to makefile
bugfix: (#18109) The restore of a hyper-v "all" backup has finished successful, but the log shows errors
new: added chksgfiles.cpp for validation mechanism of Exchange files
change: set vss_main.cpp as separate source file for building
update: added chksgfiles.hxx/lib
Bug 16206 - Validating Backup Integrity By Using CHKSGFILES - check the FDI chain against required logs by using lis, not and meta files - the new module is implemented - sm_validate.py
change: LICENSE_KEYS with license item description 'Maintenance Timeout Date'
change: (#18100) Browsing of Hyper-V cluster
new: Read path to 'rpex.exe' from registry
bugfix: Last change caclulates to high total backup size
correct cast of data size
new: 'sm_rexec' python function using new 'Popen'
new: Write RPEX command line into '.cmd' file
change: Use 'sm_rexec2' to get correct exit code on Windows
bugfix: (#18089) Kernel-mode memory leak during http restore. Select calls where replaced by curl_multi_wait()
new: 1st adoption for RPEX
new: Build Notes module if SDK is found
NEW: Update data size (DB:results.data_size) of a backup group including size of sub vSphere tasks. (Main backup saveset only includes size of config saveset, but not from VMDKs)
NEW: Allow introduce of a new medium for a data store pool, even if drive is busy (backup to another pool). This is possible because on a data store we don't need any exclusive access to a 'real' drive
new: Set 'migration_tasks.replication_type=SEP Migration' for exisiting migration tasks
NEW: Set vSphere transport options for restore too
NEW: (#18107) Allow setting local port range used for connect over http(s)
Remove unsed code
new: (#18117) Output 'Maintenance' in license overview
bugfix: vss build issue on x86
NEW: (#18117) New key MAINTENANCE_TIMEOUT_DATE
change (#18114): VM configuration files removal principle during Hyper-V backup. Now it will not be removed if LogLevel >= 4.
new: added tracing of all VSS API calls(via CVssBackupComponentDecorator)
change: Add option '-v 1|2|3...' to possible arguments
bugfix: Link with sqlite3 static library
change: Look for Notes SDK under '/usr/local/sesam/notes'
fix compiler warnings
bugfix: Link with sqlite3 static library
NEW: Create sqlite3.dll for python programs
BUGFIX: (#18082) Change open file function to not overwrite files during Notes backup. Use new function SysFileOpen() to open file read/write instead of SysFileCreate()
change: Do not use sqlite3.dll from python distribution but from common\sqlite compilation
change: Avoid duplicate trace 'Call NSFEndArchivingLogs()' in higher trace level
change: Do not set LOTUS9 always to get correct compilaation message 'LOTUS9' or 'not LOTUS9'
bugfix: (#17982) More detailed error message for Volume is not supported
bugfix: (#17956) VM backup fails with VSS_E_FLUSH_WRITES_TIMEOUT(added hint to a log)
BUGFIX: (#18083) Start all sub-restores with original drive number set by QM. If sub-restore has to use another drive, it will occupy it by submitting an placeholder job into these queue
BUGFIX: Allow '-R' and '-c' options for events executable
NEW: Compress Si3 dedup server 'INFO' log too
CHANGE: Wait 10 sec. for shutdown of Si3 dedup server. TODO: Wait 'dynamically'
change: Now Lotus 9 is set as default
NEW: SQLite version 3.8.8.3: Use journal_mode Write-Ahead Logging (WAL) to avoid 'database is locked'
NEW: SQLite version 3.8.8.3: Patch file and resource file adapted
NEW: SQLite version 3.8.8.3
bugfix: (#17949) Retry backup in case of VSS_E_WRITERERROR_NONRETRYABLE
bugfix: problem with GPT script which caused partition disappearing
change: do not attempt opening files during VFS MTF stream parse
change: inserted gitId to kernel\python files
change: inserted the gitid to the output of the sm_reformat_lis
correct trace message
BUGFIX: Read complete LIS file to find matching 'mail<->lis' items
NEW: Set minimum free disk space for Si3 dedup store to 10 GB
NEW: Start 'sm_sms_watch' in 'sm_start'
BUGFIX: Do not break during write of NOT file from a Dovecot or Cyrus backup (Error: backup undefined broken, because NOT log is incomplete after sm_reformat_lis)
new: Add function to check space on all data stores belonging to same drive group
change: Use macro 'POSIX_STAT' to use '_stati64' on Windows
BUGFIX: (#17557) Core dump of sbc during backup of VMware VM(safely close handles)
bugffix: #(17579) - RPEX - LIS File Enhancements - File Group for each mailbox database. sm_reformat_lis returned exit state 1 - failed on the DIFF and INCR backup
BUGFIX: Call 'sm_dedup_interface' always with device name as 1st parameter because if command is the 1st one parsing on Windows fails. 'sm_dedup_interface -d <device> <cmd>'
Speed up chunking in deduplication.
change: replaced the xbsa_lib.lib with the xbsa.lib to be able to build the sbc_com_interface statically again. This has to be undone in the future.
bugfix: resolving error during libbsa.dll linking due to msvcrt and libcmp fix
bugfix: (#17579) - RPEX - LIS File Enhancements - File Group for each mailbox database - changed the Python format_exchange() function in the sm_reformat_lis.py module: use the metatada file of the Exchange writer to group databases and logs
BUGFIX: (#18066) Config backup of VMware machines fails with "FindFirstFileW failed for []: WIN32 API error: 3 - The system cannot find the path specified."
CHANGE: Do not re-start replication for savesets where there is an already active task. Add migration task name to log file name
change: Use 'JOIN' to get only savesets from results belonging to correct media
bugfix: Correct parameter parsing for queue submit
bugfix: '__attribute__' not supported on Windows
BUGFIX: backint Oracle: Correct check of item saved message from 'sbc'. Backint has checked for '3108' in line, but this value could also been found in new item size message and file has been reported as saved twice. No search for '-3108'
change: Add columns to 'media_results' to save current state of a data store (filled, free...)
change: Push arguments as 'LL' list to QM
new: Macro 'InsLastLLsz(LL, char*)' to add string not defined locally to LL list (use strlen(sz)+1))
NEW: Add 'Windows 8.1'
bugfix: when installing a client on a hypercluster the core computers have to be added as clients to sesam before the actual installation process can start.
bugfix: replaced the also password column with the new os_password.
bugfix: Exchange backup is finished with error - sm_reformat_lis returned exit state 2. It occurred because of .not file format corruption for EXCHANGE. Fixed the DB_GetItemW function - setting valid status of the processed item.
bugfix: (#17951) Hyper-V backup of multi VM via Hyper Cluster resource must return error if not all cluster nodes can be accessed
Use 'replication type' instead of 'migration_cmd'
new: Write 'DB:migration_results.migration_type'
change 'migration_type' to 'replication_type'
migration_results->replication_types
new: Check for running replication tasks and update 'migration_results'
new: Set state in migration_results to 'a', even if no new migration sre started.
NEW: Return correct error, if data store isn't accessable. Set balance to '0' for not accessable data stores
move soem traces to loglevel=2
NEW: Improved Sesam SAP HANA install scripte
do write to DB:result_lbls, because INIT isn't written to results any more
change: only check for the user name and password in the db (os_user, os_password) and install, update, deinstall if they are present. But don't pass the user name and password to the sm_remote_installer since the sm_remote_installer takes them from the db anyways if they are present.
change: If install sequence was chosen and sesam is already installed on the system then an update is executed instead.
BUGFIX: (#16206) - Validating Backup Integrity By Using CHKSGFILES. The first phase of implementation is done - adapting the client side - print out "required" logs for database in "Dirty Shutdown" state.
bugfix: (#17833)Backup of HSM files: HSM Reparse points should be followed with sbc option '-o follow' (added additional log messages in case of HSM)
add 'DB:migration_results.migration_type'
new: Allow mount of savesets on SI3 dedup store Windows
change: slu - when building topology examine if serial number is substring of device identifier
bugfix: (#18059) HyperV SMB restore does not work
add log
bugfix: Set source pool for migration in case copy jobs are submitted ti queue. TODO: Unify command line creation
new: Check time range for saveset replication
new: Forward time range for saveset replication to 'sm_replicate'
'replication_type' instead of 'migration_type'
bugfix: fixed Hyper-V/Exchange adding backup components
change: instead of streaming the msi packages two times the chained installation will use the msi files located in the skel directory during installation and update. This reduces the size of the installation packages.
bugfix: (#16181) fixed some minor issues concerning the setting of the user name and password of the sm_main service.
bugfix: (#18028) Hyper-V restore fails with: Error while unpacking data. Bad archive
change: inserted the creation of the salt key to the installation and to the update of the sesam server (set_salt). Also fixed a bug in the salt key creation. Erased the activate_uac since it is no longer needed.
bugfix: (#18052) - [BackupProcessing: For the specified backup source no volumes could be determined. Sources - \Microsoft Exchange Writer]. The backup source type is set to value "WRITERS" to get able to create snapshot.
new: Set LD_LIBRARY_PATH to '/usr/local/sesam/lib' for 'cxfreeze'
new: extended DISKPART script generator to support GPT disks.
bugfix: sbc core due to stack overflow fixed. There was infinite recursion with trace()->vttrace()->mbstowcsl()->trace()
new: replaced the O&O version 9.0.252 with the current one 9.0.261
change: adjusted vmware default buffer size on linux(default=128)
bugfix: use correct buffer size in vmware(linux)
change: Makefile updated to fix missing ssl library reference
change: makefiles updated to use new curl version
bugfix: Check for NULL pointer and empty string
bugfix: added additional logs to regarding vmware buffers
BUGFIX: (#18027) - Slow browsing of MS-Exchange server. Request only MS Exchange writer metadata instead ones of all available writers in the system - it allows to get Exchange items in 3-4 seconds.
change: inserted closing trace line.
BUGFIX: (#18040) A vSphere restore from a copy backup sporadically fails since sesam V 4.4.1.46(release of pipe resourcing)
bugfix: call IVssBackupComponents::BackupCompete in case when at least one writer/component is involved into backup process. Message: CVssServer::CreateSnapshot(): IVssBackupComponents::BackupComplete] - VSS_E_BAD_STATE
change: set ctime and mtime to now if not specified; parse create time and modification time correctly
change: Remove unused DISK_HARD based test backup during Sesam installation functions
change: use mtime for LastAccessTime
BUGFIX: (#18040) A vSphere restore from a copy backup sporadically fails since sesam V 4.4.1.46
change: sm_vfs attempts several locales when extracting file information from LIS and CPIO
change: mbstowcsl accepts list of locales to attempt before giving up
change: added filesystem testing utility
change: read only up to 64k from each saveset when locating relative file position
change: handle uninitialized cache in cache destroy
new: (#18031) inserted the activate_uac switch for the activation of the user access control in sm_setup
log and source format change only
BUGFIX: Quote file name of Si3 fsck dump file (duirng read), so it works on pathes with space
bugfix: Fixed CSV backup of VM's on different mount points
change: from now on the client user name and password will be obtained from the columns os_user and os_password from the clients table.
bugfix: DB export/import: Call 'popen' on UNIX with 'r' instead of 'rb', because not supported here. Causes core dump of 'sm_db' during update
change: Hyper-V relocation message text is fixed.
bugfix: Correct check if replication type already exists
new: Call 'sm_client dir <mnt>' on RDS where saveset is mounted
bugfix: fixed syntax error.
bugfix: the var and bin dir could not be set using the command line.
bugfix: pywintypes34.dll should only be inserted for x64 not the other way around.
change: exclude the pywintypes34.dll from x86 kits.
change: Separate mount points by ', ' instead of ' '
BUGFIX: Quote file name of Si3 fsck dump file, so it works on pathes with space. (file C:/progam is created)
change: inserted the pywintypes34.dll to the x64 server, client and gui kit since it is needed for the sm_ssh, sm_sshd only for x64.
new: insert the pywintypes34.dll to the client and gui packages. It is required by cxfreeze.
change: inserted the x86 specific productcode to the sesam installer.
change: resolving memory leaks reported by valgrind
change: clear cache on exit
change: (#17992) install O&O also if sesam has already been installed and execute an update of O&O if it has already been installed on the system.
new: added functionality to XML parser to find GPT disk attributes
bugfix: added additional vss traces
change: (#17992) install O&O also if sesam has already been installed and execute an update of O&O if it has already been installed on the system.
bugfix: Resolving missing CRYPT_LIB reference
bugfix: Resolving missing curl dependencies
bugfix: resolving missing CRYPT library location
bugfix: (#18015) Insert retry in the Hyper-V browsing(fixed sbc root cause)
change: inserted a warning in case the product code of O&O changed. The build will execute in that case.
change: Avoid compile warning C4552: '!' : operator has no effect
BUGFIX: (#17983) sm_lis_items doesn't show all backed-up volumes. The volume F: is labeld 'Staff/Home-Prog'. Set level back to 0 if matching length is 1.
change: Trace log level in print_STAN() revised
change: close file handle on return
BUGFIX: (#18000) Backup of HSM files: Check logic for HSM Reparse Points specific Reparse Tags HSMxxx. Backup reparse tag NTRP stream for Microsoft symlink and mount point. Other Microsoft tags are backed up with standard file data stream. In case of Non-Microsoft reparse tags only IO_REPARSE_TAG_HSAG_HSM (Metalogix) will be backed up without warnings.
bugfix:(#18018) Changes in makefiles related to libcurl version change
bugfix: (#18018) BSAHttpEndData: NEGATIVE reply: (0) issue fixed. Win32 select(0, ...)  call was replaced by Sleep
fix compiler warning
bugfix: fixed grammatical error
change: _wcsdup to wcsdup
change: fix warnings
change: sm_vfs resolve memory leaks
change: log entry reworked
change: sm_vfs performance enhancement - store previous search
change: sm_vfs win - cast to expected type
change: set size of entry to 0 if not file type
change: print more info in cache dump
change: removed obsolete code
change: potential crash resolved
BUGFIX: (#17943) Restore of specific files located on Windows Dedup volume fails with 'no valid MTF Stream ID was found'. The 'offset in the file' value must be reduced from the total file size.
bugfix: build issue with sm_sms_backup
bugfix: (#18015) Insert retry in the Hyper-V browsing
change: (#17957) Concurrent VSS-backup lock control algorithm redesign.
bugfix: Core dump if SQL no 'FROM' keyword in SQL statement
BUGFIX: (#17943) Restore of specific files located on Windows Dedup volume fails with 'no valid MTF Stream ID was found'. The 'offset in the file' value 8 Bytes must be skipped from the SPAR stream in the end.
change: (#17957) Concurrent VSS-backup lock control algorithm redesign.
change: Use makefile.wcommon to get $(COMMON_WINDOWS_DIR)
NEW: Set 'results.i_name' during 1st insert into DB, so even task group entries have this value set
BUGFIX: (#17320) MSSQL Backup of database with special characters fails. Use wide char to pass database name to SQL API
change: sm_vfs skip intended number of bytes
fix compiler warnings
change: sm_vfs unlock mutext instead of locking it
change: build dedup libdependency on windows correctly
change: sm_vfs protect seek&read with mutex
change: sm_vfs rework si3 handling
change: (#17957) Trace message 'Unlocked concurrent VSS backup processes' must appear without setting higher log level
Concurrent VSS-backup lock control algorithm redesign.
BUGFIX: Getting stpd port for '-C <sesam server': If no DB entry is found use default port. TODO: Check, if client name with 'client_id=0' has the same value then 'gv_server' variable in 'sm.ini'
new: (#18019/#18020) New license keys MSSharePoint_Extension_GB, SEP_DeDup_REP. Obsolete license keys removed: FDS_BASIS_TB, FDS_EXTENSION_TB, MARATHON_EVERRUN, MARATHON_EVERRUN_SEATS
bugfix: added ability to backup VM's on different SMB volumes
bugfix: (#17957) Do not lock execution wiht old synchronize method. Now only the Hyper-V Writer state is traced.
NEW: Log with ISO date/time format like all other Sesam modules
change: Moved the trace messages about semaphore locking from debug to VSS_Trace level
NEW: Add pid to backint log file to avoid huge files
bugfix: Line for VOLB was printed without space between offset1 and offset2 in case of MSQL (MQDA, MSCI Block or MQCI stream)
change: BLOW_UP_SLL() Do n ot add parent directory if sel line is a 'd'irectory line
NEW: Print out and write Git ID into backint log
new: Update 'based_on' and 'based_on_full' for sub tasks (vSphere...) with values from main task
bugfix: Return with error if 'sm_vfs' isn't executing after 2 sec. any more
Concurrent Hyper-V backup lock control log records has been added.
Concurrent Hyper-V backup lock control log records has been added
BUGFIX: (#18012) Restore ends with 'submit external job to restore in saveset' but no external savesets are used. A directroy of a previous INCR saveset appeared in ssl file. If no items are selected in a INCR saveset then the memorized parent directory must be reset before processing the next INCR saveset
bugfix: incorrect context in Hyper-V backup
change: (#17992) use reduced user interaction for the first O&O installation tests.
bugfix: Call OpenLog() before ini_read_thread_values() to allow trace messages
change: include the new version of O&O version 9.0.252
change: Function renamed to i_FULL_DIFF_INCR_CHAIN(...)
change: Saveset length set to 64 - e.g. based_on_full VARCHAR(64)
 Bug 17954 localization by debug_trace mechanism.
Revert "change: reverted the changed made for including the O&O msi installer in the build routine."
BUGFIX: Set option 'Descend directories on other file systems' for task SESAM_BACKUP to back up data under 'var' reparse point
change: use the newest O&O msi package version 9.0.251.
bugfix: (#17690) Hyper-V backup from a cluster fails because of VSS_E_WRITERERROR_TIMEOUT(added information about clustered machines)
bugfix: fixed multiple VM's bkp
change: windows compile fix
change: reverted the changed made for including the O&O msi installer in the build routine.
change: sm_vfs enable thread id trace
change: enhance trace method - enable thread id print
change: another Trace messages level revisit
change: skip bytes in chunks if bytes to skip bigger than size_t
changed: enhanced error reporting in case of skip_bytes
change: sm_vfs fixed a potential multithread issue
change: introduce cache mutex on linux
change: check if tmp file created
bugfix: Windows: Do not stop reading import file if Ctrl-Z is contained in the import file
BUGFIX: (#17997) - Handle the exclude keyword -x 'all_writers' to specifies an auto-release, nonpersistent shadow copy created without writer involvement
change: Trace messages level revised again
change: Trace messages level revised
change: sm_vfs accept only f and d, ignore others
change: sm_vfs/si3 reopen dedup file handle if EOF hit
bugfix: memory correctly allocated
change: i2dedup client free memory when handle closed
change: Use 'sm_vfs_{pid}_...' log file names
Bug 17954 localization by debug_trace mechanism.
change: tried to make some of the log/error messages clearer.
* change: Add DbgHelp.lib to link rule of xbsa.dll to be able to build debug version (via RT_LIB)
change: sm_vfs always contains debug information
change: enable line reporting and variable reporting on crash
new: If no additional arguments are set for a follow up wvent then append no empty argument, because CLI argument parsers throws an exception in version 4.4.2
new: If no additional arguments are set for a follow up wvent (migration) then append no empty argument, because CLI argument parsers throws an exception in version 4.4.2
bugfix: fixed multiple VM's backup
Bug 17954 localization by debug_trace mechanism.
bugfix: STPD HTTP synchronization code changed from mutexes to semaphores. STPD HTTP stability enhanced
change: adjusted the path to the Xcode folder, which had changed due to the xcode update to version 5.
bugfix: (#17978) - T-SQL restore is failed with using HTTP and FTP interfaces. The core dump is detected and fixed for T-SQL restore operation.
BUGFIX: (#17975) In function 'saveset_tree': Return saveset information in case of migrated saveset given, but original one doesn't exist any more. TODO: Return with warning, because this is an unexpected warning.
bugfix: (#17992) use the stream file function instead of the run-time path function for including the O&O msi file.
bugfix: Set '-d' switch for Si3 mount only
remove code for OS not supprted any more
bugfix: (#17690) Hyper-V backup from a cluster fails because of VSS_E_WRITERERROR_TIMEOUT
bugfix: (#17989) MS SQL Server restore failed with: 'E012-RESTORE Restore was not successful. Error: DB Module: []'. In recent implementation the free_pipe_get_status() was called twice.
change: (#17992) included the installer of O&O in our own installer using chained msi packages function. For this purpose the msi 4.5 is required which will also be distributed during installation as redistributable.
BUGFIX: (#17990) If client is accessed via CTRLD then backup log files (not files) and restore log files are displayed with extra linefeed. Do not write the buffer received via stpd with operating system specific line conversion but write single lines.
Bug 17954 localization by debug_trace mechanism.
bugfix: (#17980) For client/rds/gui with client also changed the remove and install to only install the sm_main service which takes care of updating the executable of the service..
bugfix: (#17980) During an udpate of a sesam kit the sm_main service is not removed and installed but rather only the service executable is updated to ensure that the user and password will not get lost.
Bug 17954 localization by debug_trace mechanism.
NEW: Script to adopt SEPsesam client installtion to run on a Fujitsu FlexFrame node
Cleaning enabled
Python 3 version build added
bugfix: (#17833)  Backup of HSM files: HSM Reparse points should be followed with sbc option '-o follow' (Added logs and All flag for both options)
change: Added trace() function
bugfix: (#17833) Backup of HSM files: HSM Reparse points should be followed with sbc option '-o follow' (FILE_FLAG_OPEN_NO_RECALL)
change: Set default incativity timeout to 300s (5 minutes) and allow to set values until 7200s (2 hours)
change: Add DbgHelp.lib to link rules (via RT_LIB)
change: for macos replace echo with printf since echo on mac does not have an -e switch.
bugfix: Try to mount VMDKs from successful backups only. Failed ones will be omitted now
bugfix: separate case for I_ACT_END_PROCESSING condition. To stop saveset restore process on demand. bugfix: ASR Writer backup. No volumes was added to snapshot.
change: Add DbgHelp.lib to link rule of sm_user_manager.exe and sm_dedup_interface.exe (via RT_LIB)
BUGFIX: (#17854) Adjust buffer in case of VMware (SAN) backup
Hyper-V - ordinary, SMB and CSV backups are works (at least on my test VMs).
Bug 17954 localization by debug_trace mechanism.
change: Error message revised
bugfix: erase the trailing quotes from the java path in the path variable.
Bug 17954 localization by debug_trace mechanism.
bugfix: Last commit e7d41b064 lead to empty version.h
Bug 17954 localization by debug_trace mechanism.
BUGFIX: (#17944) - Incremental SQL backup has to stop if another backup is still active. Add "version.bat" file.
change: changed condition to detect BAREMETAL recovery.
BUGFIX: (#17944) - Incremental SQL backup has to stop if another backup is still active. Use kernel synchronization objects for prevent parallel backup of the same sql database.
new: Support different replication (migration) types
new: Check for valid backup type and exit with error message instead of an exception if it's unknown
new: Add 'gv_ro_gui' path to 'sm.ini' during GUI installation too
Bug 17954 localization by debug_trace mechanism.
change: Messages for E011_BACKUP_MISSING_FULL and XX_SBC_COM_ERR_1026 revised
Bug 17954 localization by debug_trace mechanism.
change: (#17964) If saveset chain cannot be retrieved then sbc_com_interface must return a specific error message
bugfix: (#17742) Exclude for file download for VADP backup (supported regexp)
Bug 17954 localization by debug_trace mechanism.
BUGFIX: (#17964) Sesam does not update retention time of VMware CBT base saveset in case of INC of DIFF backup. Now the session_id is used to retrieve the basic FULL raw vmdk backup
Bug 17954 localization by debug_trace mechanism.
VSS.dll debug_trace mechanism integration fixed.
bugfix: In case of sparse file with size 0 the calculation must take care about the size of the offset indicator
Additional logs are inserted to be able to fix bug 17954.
bugfix: Since commit 38ddba5 'bugfix: Restore with small sparse files ends with: Error while unpacking data' the following offset to the SPAD stream was 8 Bytes to big
bugfix: Since commit 9f0a6d3 '(#10043) Additional restore options' directories and files are only restored with option '-o overwrite'. The function GetFileAttributesEx() returns a nonzero value if file exists
change: remove the delete statements of the vddk and the O&O.
change: copy don't move the msi and the VDDK files from the new location.
change: corrected the name of the msi folder for O&O.
Compiler warnings and memory leak in 'vss_sqlmgr.cpp' are fixed.
BUGFIX: (#17957) (#17958) Avoid parallel execution of Hyper-V backup.
change: Use additional description '- QueueManager:runjob()' in daily log message.
BUGFIX: (#17960) Windows: sm_qm_main ends execution if follow_up executable does not exist. Go on with execution.
change: dedup client - handle feof correctly
change: fix makefile on sms_interface to react on changed files
bugfix: (#17265) Backup over files system bounderies not possible on Windows
Bug 17954 localization by debug_trace mechanism.
bugfix: sm_dedup_interface handle get correctly
remove: deleted the VDDK zip files. They will be located in a separate repository in the future since they are too big.
remove: deleted the O&O installer packages. They will be located in a separate repository in the future since they are too big.
change: move the VDDK and the OO installer packages to the sesamdev folder.
 Hyper-V access and related modules - intermediate file has been removed.
Hyper-V access and related module update. VHD mount feature added Microsoft VSS-support include files are removed from Sesam sources, so original ones (from current compiler) will be used instead.
bugfix: corrected the wrong skel directory for the copy of the VMware and O&O file.
bugfix: inserted a variable for the VDDK and O&O version and corrected the copy of the O&O msi package for the gui installer package.
new: Added an install and an uninstall function for the O&O software. Installation will be executed during Installation of sesam and uninstall will be executed during uninstall of sesam.
new: added the copy of the o&o msi installer to the windows installation package.
new: initial checkin of the OO msi installer